In the previous chapter, we examined the strong biblical case for the universal scope of the atonement — the teaching that Christ died for all people without exception. We walked through passage after passage, from John 3:16 to 1 John 2:2 to Hebrews 2:9, and we saw that the New Testament consistently presents the death of Christ as intended for the entire human race. The evidence, I believe, is overwhelming.
But there is another side to this debate, and fairness demands that we hear it out. The Calvinist doctrine of limited atonement — also known as particular redemption or definite atonement — teaches that Christ died only for the elect, that is, only for those whom God chose before the foundation of the world to save. On this view, the cross was never intended to provide atonement for those who would ultimately be lost. Jesus bore the sins of the elect and the elect alone.
This doctrine is the "L" in the well-known TULIP acrostic, which summarizes the "five points of Calvinism": Total depravity, Unconditional election, Limited atonement, Irresistible grace, and Perseverance of the saints. Limited atonement is often considered the most controversial of the five points — and for good reason. Even many within the Reformed tradition have rejected it while affirming the other four points.1
This chapter is devoted to a careful, systematic response to the arguments for limited atonement. My thesis is straightforward: the Calvinist doctrine of limited atonement is exegetically, theologically, and historically untenable and should be rejected in favor of the universal scope of Christ's atoning work. I want to be fair to the strongest versions of the argument. Many brilliant and godly Christians have held this position, and I have no desire to caricature their views. But I am convinced that when we examine the evidence carefully, the case for limited atonement does not hold up.
Chapter Thesis: The Calvinist doctrine of limited atonement (particular redemption) — that Christ died only for the elect and not for all people — is exegetically, theologically, and historically untenable, and should be rejected in favor of the universal scope of Christ's atoning work.
Before responding to limited atonement, we must first understand it. I believe in steel-manning the opposition — presenting their case in its strongest form before engaging it. Anything less would be intellectually dishonest. So let me lay out the main arguments that Calvinists offer in defense of particular redemption.
The most powerful argument for limited atonement is not really an exegetical one. It is a logical and theological one. It goes like this: If God unconditionally elects certain individuals to salvation (the "U" in TULIP), and if the grace that brings them to faith is irresistible (the "I"), then it seems to follow that Christ must have died specifically and only for those individuals. Why would Christ die for people whom the Father never intended to save? If God knew from eternity past exactly who would be saved, and if He sovereignly determined the outcome, then surely He would have sent His Son to die specifically for those chosen ones — and not for those whom He had passed over.
This is a compelling piece of reasoning, and I want to acknowledge its force. As Roger Nicole, a respected Reformed scholar, has helpfully clarified, the question is not about the value of Christ's death. Virtually all Calvinists agree that Christ's death has infinite value — more than enough to atone for the sins of the whole world and a thousand worlds besides. The question, rather, is about the design or intention of the atonement. For whom did Christ intend to die? Who was on His heart when He went to the cross?2
We should also note that many defenders of limited atonement prefer the term "definite atonement" or "particular redemption" precisely because "limited" sounds like they are diminishing the cross. They are not saying Christ's death was limited in its worth. They are saying it was limited — or more precisely, definite — in its intention. Christ had a specific people in mind when He died, and His death infallibly accomplishes their salvation.
Calvinists point to several passages in the New Testament that seem to present the atonement in particular rather than universal terms:
John 10:11, 15: "I am the good shepherd. The good shepherd lays down his life for the sheep." Jesus does not say "for the whole world" here. He says "for the sheep." And in verse 26, He tells certain people, "you are not my sheep." If Jesus lays down His life for the sheep, and some people are explicitly not His sheep, does this not imply that He did not die for those who are not His sheep?
Ephesians 5:25: "Christ loved the church and gave himself up for her." Paul says Christ gave Himself for the church — not for humanity in general. Does this not indicate that the intended beneficiaries of the cross are the church and the church alone?
Matthew 1:21: "He will save his people from their sins." The angel announces that Jesus will save "his people" — a definite, particular group.
Acts 20:28: Paul tells the Ephesian elders to care for "the church of God, which he obtained with his own blood." Again, the blood of Christ is linked specifically to the church.
Romans 8:32–34: "He who did not spare his own Son but gave him up for us all, how will he not also with him graciously give us all things?" Paul's logic here seems to be: God gave His Son for "us" (believers), and because He did this greatest thing, He will surely do the lesser thing and give us everything else we need. The argument appears to assume that all those for whom Christ died will certainly be saved and receive all things. If Christ died for all people without exception, this argument seems to prove universalism — that all will be saved — which no orthodox Christian accepts.
John 17:9: "I am not praying for the world but for those whom you have given me." In His high priestly prayer, Jesus explicitly says He is not praying for the world. If Christ intercedes only for the elect, does this not suggest He died only for the elect?
Beyond these texts, Calvinists raise several theological arguments:
The "wasted atonement" argument: If Christ bore the punishment for the sins of all people, but many of those people will end up in hell, then Christ's atoning work was "wasted" on them. How can it be just for God to punish sin twice — once on Christ and once on the sinner? Did Christ really suffer for sins that will never be forgiven?
The Trinitarian unity argument: If the Father elects only some, and the Spirit regenerates only some, but the Son dies for all, then we have a kind of discord within the Trinity. The Father and the Spirit are working toward one goal (saving the elect), while the Son is working toward a broader goal (dying for all). Should we not expect the three Persons to act in perfect harmony?
The efficacy argument: If Christ's death merely makes salvation possible but does not actually accomplish it, have we not cheapened the cross? Is it not more glorious to say that the cross actually saves, rather than merely opening a door that people may or may not walk through? As J. I. Packer argued, does universal atonement not "limit" the cross more drastically than Calvinism does — by denying that Christ's death, as such, is sufficient to save anyone?3
I have tried to present these arguments as clearly and forcefully as I can. They are not trivial. They come from careful, thoughtful theologians who love the Scriptures and the gospel. But I believe every one of them can be answered. Let us work through them systematically.
Key Distinction: The question in the limited atonement debate is not about the value of Christ's death (all sides agree it has infinite value) but about the design or intention of the atonement — for whom was it intended?
Let us begin with the biblical texts that Calvinists cite in favor of limited atonement. As David Allen has pointed out with great force, the first and most important thing to notice about every single one of these texts is that not a single one says Christ died only for the elect. This is the crucial point. There is no statement anywhere in Scripture that says Jesus died only for the sins of a particular group. Limited atonement, as Allen rightly notes, is "a doctrine in search of a text."4 The arguments for it are almost entirely logical and deductive rather than exegetical.
When Jesus says, "I lay down my life for the sheep" (John 10:11), He is making a positive statement about His love for His own. He is not making a negative statement that excludes everyone else. This is what logicians call the negative inference fallacy — assuming that because a proposition affirms something about one group, it necessarily denies it about all other groups.
Think about it this way. If I say, "I would give my life for my children," no one would conclude that I am saying, "and I would let everyone else die." My statement expresses special love, not exclusive love. When Paul writes, "Christ loved the church and gave himself up for her" (Eph. 5:25), he is describing Christ's particular love for His bride. But this says nothing — absolutely nothing — about whether Christ also died for those outside the church. Paul himself, in the very same letter, can speak of Christ's love for the church and affirm that God desires "all people to be saved" (1 Tim. 2:4, written to the same general audience).
Consider Paul's personal testimony in Galatians 2:20: "The Son of God, who loved me and gave himself for me." Did Paul mean that Christ died only for him? Of course not. He is expressing a deeply personal truth — Christ's death is for me — without denying its broader scope. The same logic applies to every "particular" text. Saying that Christ died for the sheep, the church, or His people does not exclude others any more than Paul's "for me" excludes the rest of humanity.5
What about John 10:26, where Jesus tells certain people, "you are not my sheep"? Does this not imply a fixed group for whom Christ died and another group for whom He did not? We need to read the passage carefully. Jesus is speaking to the unbelieving Jewish leaders who have rejected Him. His point is that their unbelief demonstrates that they are not currently among His followers — they are not listening to His voice. But this does not prove that Christ's death was not offered for them. Throughout John's Gospel, unbelief is consistently presented as a culpable refusal of what God has genuinely offered (John 3:18–19; 5:40). People are not lost because the atonement was insufficient for them or because Christ did not die for them. They are lost because they have rejected the gift.
Moreover, John 10 must be read alongside John's other statements about the scope of Christ's work. John 1:29: "Behold, the Lamb of God, who takes away the sin of the world." John 3:16: "For God so loved the world, that he gave his only Son, that whoever believes in him should not perish." And the definitive statement in 1 John 2:2: "He is the propitiation for our sins, and not for ours only but also for the sins of the whole world." How can we take "the sheep" in John 10 as a limitation on the atonement when the same author says Christ is the propitiation for "the whole world"?
The argument from Romans 8:32 is more subtle and deserves careful treatment. Paul's logic does seem to be: God gave His Son for "us," so He will certainly give "us" all things. Does this not mean that all for whom Christ died will inevitably be saved?
Yes — but who is "us" in this context? Paul is writing to believers. His chain of assurance in Romans 8:28–39 is addressed to those who are already in Christ by faith. Paul is saying: You who believe, you who are in Christ — take heart! The God who did not spare His own Son for you will not abandon you now. This is a magnificent statement of assurance for believers. It is not a statement about the intended scope of the atonement.
Think of it this way. If a father pays the full college tuition for all of his children, and one child never enrolls, the father can still say to the child who did enroll: "I paid for your entire education — do you think I won't also buy your textbooks?" This reasoning holds perfectly for the enrolled child, even though the tuition was also paid for the child who never showed up. Paul's argument in Romans 8:32 works because believers have received the benefit of Christ's death through faith. It does not require that Christ died only for them.
What about Jesus' statement, "I am not praying for the world but for those whom you have given me" (John 17:9)? Does this not prove that Jesus had only the elect in mind?
We need to be very careful here. First, the context makes clear that Jesus is praying specifically for His disciples — the particular men who were with Him that evening. He is praying for their protection, their unity, and their sanctification as they face the coming crisis. There would be no point in praying these specific things for the unconverted world, because these prayers concern the specific needs of His followers.
But here is what is truly significant: Jesus does not stop at verse 9. In verse 20, He says, "I do not ask for these only, but also for those who will believe in me through their word." And then — crucially — in verses 21 and 23, He prays "so that the world may believe that you have sent me" and "so that the world may know that you sent me and loved them." As Allen and others have pointed out, Jesus is actually praying for the world's salvation in this very chapter. He prays that the unity of believers will have the result of bringing the world to faith. The word "world" here cannot be limited to the elect — it refers to the unbelieving world that Jesus desires to reach.6
Harold Dekker, formerly of Calvin Theological Seminary, argued this point persuasively. When Jesus says He does not pray "for the world" in verse 9, He means that these particular petitions (for protection, unity, sanctification) only apply to those who already belong to Him. But this says nothing about whether He died for the world or desires the world's salvation. In fact, John 17:21 and 23 show that He does.7
There is a further logical point that is often overlooked. Defenders of limited atonement argue: If Christ prays only for the elect, He must have died only for the elect. But this logic can be turned on its head. All would agree that if Christ prays for a person, He must have died for that person. John 17:21 and 23 clearly show Christ praying for the world — that the world may believe and know. Therefore, He must have died for the world.8
When we place the "particular" texts alongside the "universal" texts, the comparison is decisive. The particular texts say Christ died for "the sheep," "the church," "his people." The universal texts say Christ died for "all" (1 Tim. 2:6), "everyone" (Heb. 2:9), "the world" (John 1:29; 3:16), and "the whole world" (1 John 2:2). The particular texts never say "only." The universal texts explicitly extend the scope beyond any limited group. As we argued extensively in Chapter 30, the natural reading of these texts is genuinely universal — all people without exception.
To make the particular texts teach limited atonement, Calvinists must add the word "only" — Christ died only for the sheep, only for the church. But that word is not in the text. To make the universal texts consistent with limited atonement, Calvinists must restrict their meaning — "world" means "the elect from every nation," or "all" means "all kinds of people, not all people without exception." But as we demonstrated in Chapter 30, these restrictive readings are strained and contextually implausible, especially in passages like 1 John 2:2 where "the whole world" is explicitly contrasted with believers.9
The Negative Inference Fallacy: Saying "Christ died for the sheep" does not mean "Christ did not die for anyone else" — just as saying "I love my children" does not mean "I do not love anyone else." The particular texts express special love, not exclusive love. Not a single text in Scripture says Christ died only for the elect.
We come now to what is perhaps the most famous and frequently deployed argument for limited atonement. It is often called the "double payment" or "double jeopardy" argument, and it runs like this:
If Christ bore the punishment for a person's sins on the cross, and that person later goes to hell, then God is punishing the same sins twice — once on Christ and once on the sinner. This would be unjust. Therefore, either Christ died only for the elect (who will certainly be saved), or universalism is true (all are saved). Since universalism is false, limited atonement must be true.
John Owen, the great Puritan theologian, put this argument in its most famous form. In The Death of Death in the Death of Christ, he wrote: "If the full debt of all be paid to the utmost extent of the obligation, how comes it to pass that so many are shut up in prison to eternity, never freed from their debts?"10
This is a powerful argument, and I understand why many find it persuasive. But it contains a fatal flaw: it rests on a commercial understanding of the atonement that the New Testament does not support. Let me explain.
The double payment argument treats Christ's death as though it were a financial payment — like paying off a credit card debt. If the debt is paid, then it is paid, and no one can demand it again. If I owe the bank $10,000 and someone pays it for me, the bank cannot come back and demand the money from me as well. The debt is settled.
But this is precisely the mistake. The language of "debt" and "ransom" and "redemption" in the New Testament is metaphorical. It draws an analogy between what Christ accomplished and financial transactions, but it does not mean the atonement operates on the same mechanism as a bank payment. As Allen helpfully illustrates, there is a crucial difference between commercial debt and criminal debt.11
Imagine you and a friend are eating at a restaurant. When the bill arrives, you realize you have no money, so your friend kindly pays for both of you. The restaurant owner does not care who pays — the bill is settled. This is commercial debt. Now imagine that after the meal, you rob the restaurant and steal $500. Your friend, out of kindness, repays the stolen money. Later, when you are caught, can you say, "You can't arrest me — someone already paid for my crime"? Of course not. Criminal debt does not work that way. The penalty for a crime is not automatically discharged just because someone else offered to bear the consequences.12
Sin is not a commercial debt — it is a crime against God's moral law. The atonement is not a payment into a cosmic cash register. Christ's death provides the ground or basis on which God can justly forgive sinners, but it does not automatically and mechanically discharge the guilt of every person for whom it was made. There is a condition: faith. Just as a pardon issued by a governor must be received by the prisoner to take effect, so the benefits of Christ's atoning death must be received by faith. William Lane Craig makes this point clearly — the atonement accomplishes our potential redemption, which is actualized in individual lives only as people come to be united with Christ through repentance and faith.13
There is another problem with this argument that is often overlooked. If the double payment logic is correct — if Christ's death automatically and necessarily secures the salvation of every person for whom He died — then why are the elect not justified at the moment of the cross? Why are they born as "children of wrath" (Eph. 2:3) and live for years in unbelief before they come to faith? On the double payment scheme, the debt was paid at Calvary. Why does God wait to apply the benefit?
As Charles Hodge — himself a Calvinist — recognized, the commercial model cannot be right: "There is no grace in accepting a pecuniary satisfaction. It cannot be refused. It ipso facto liberates. The moment the debt is paid the debtor is free; and that without any condition. Nothing of this is true in the case of judicial satisfaction."14 If Christ's death works like paying off a loan, then salvation is owed to the elect, and faith is irrelevant. But Scripture consistently presents faith as a genuine condition for receiving the benefits of the atonement (John 3:16; Acts 16:31; Rom. 10:9–10).
The truth is that Christ died one death — the death that all sinners deserve under the law. In bearing the curse of the law as defined by the law, He provided a sufficient basis for the salvation of every human being. As Hodge himself wrote, "What was suitable for one was suitable for all. The righteousness of Christ, the merit of his obedience and death, is needed for justification by each individual of our race, and therefore is needed by all."15
Allen raises a further point that I find quite compelling. In 1 John 2:2, John writes that Christ "is the propitiation (Greek: hilasmos, ἱλασμός) for our sins, and not for ours only but also for the sins of the whole world." Advocates of limited atonement sometimes argue: if Christ is the propitiation for the whole world, then the whole world's sins are propitiated, and the whole world must be saved. Since the whole world is not saved, "the whole world" must mean something less than all people.
But this argument depends on an illegitimate conversion of a noun into a verb. "Propitiation" (hilasmos) is a noun. It describes what Christ is — the one who provides the means of atonement. It does not describe an already-completed action with automatic results. Compare 1 John 2:1, where John says Christ "is" our "Advocate" (paraklētos, παράκλητος). This does not mean Christ has already advocated for every person regardless of their response. It means He is the Advocate to whom sinners may go. Similarly, Christ is the propitiation — the means by which sinners may find forgiveness — for the whole world. But propitiation accomplished does not mean propitiation automatically applied. Without faith, the benefit remains unapplied.16
The Commercial Fallacy: The double payment argument treats the atonement like a commercial transaction — if the debt is paid, the debtor goes free automatically. But sin is a criminal matter, not a commercial one. Christ's death provides the basis for forgiveness, but faith is the condition for its application. Propitiation accomplished does not mean propitiation applied.
John Owen's "trilemma" is another classic argument for limited atonement, and it is closely related to the double payment argument. Owen argued that Christ died for either (1) all the sins of all people, (2) all the sins of some people, or (3) some sins of all people. Options 1 and 3 are problematic, he claimed. If option 1 is true, and Christ died for all the sins of all people — including the sin of unbelief — then how can anyone go to hell for unbelief? The sin of unbelief has already been atoned for. Therefore, either everyone is saved (universalism) or option 2 must be correct: Christ died for all the sins of some people (the elect).17
This argument has been enormously influential, but it faces several serious problems.
If Owen's logic is correct, and Christ died for all the sins of the elect — including their sin of unbelief — then why are the elect not saved at the cross? Why do they spend years living in unbelief before coming to faith? If unbelief has already been atoned for, then the elect should be born saved. But they are not. Paul says that even the elect were once "children of wrath, like the rest of mankind" (Eph. 2:3).
If Owen responds that the benefits of the cross are not applied until the person comes to faith, then he has made exactly the same distinction that defenders of unlimited atonement make: the atonement is accomplished for a group, but it is applied only when individuals believe. The only question is how large that group is. Owen's trilemma cannot decide this question, because it proves too much on his own terms — it would mean the elect are spared God's wrath whether they believe or not, which contradicts Ephesians 2:3.18
Owen treated unbelief as simply one sin among many — a specific offense that Christ either did or did not atone for. But as Neil Chambers has argued, this is a form of "polemical reductionism." Unbelief is not just an individual act of sin. It is also a state — the state of being unregenerate, of not yet having come to saving faith. The cross deals with the guilt of sin directly. But the change from an unregenerate state to a regenerate state — from unbelief to faith — is accomplished not directly by the cross but indirectly, through the preaching of the gospel and the regenerating work of the Holy Spirit. Owen's argument collapses these two different things into one.19
Perhaps the most fundamental problem with Owen's trilemma is that it assumes a quantitative understanding of imputation — as if Christ was loaded up with so many individual sins, like "sin-bits," and each one was checked off a cosmic ledger. But this is not how imputation works in Scripture. Just as believers are not imputed with specific individual acts of Christ's righteousness — so many units of "goodness" — but with righteousness categorically, so also Christ was not imputed with a tallied list of specific sinful acts. He was treated as though He were sinful. He bore the curse of the law in a comprehensive, categorical way — not as a quantified sum of individual transgressions.20
Christ died one death — the death that the law demands of every sinner. In paying what one sinner deserves, He paid what every sinner deserves. The atonement is not a matter of divine arithmetic; it is a matter of divine justice and divine love.
Another argument for limited atonement appeals to the unity of the Trinity. If the Father elects only some, and the Spirit regenerates only some, then surely the Son must have died only for some. Otherwise, there is a kind of "discord" within the Godhead — the Son working at cross-purposes with the Father and the Spirit.
This argument sounds impressive, but it rests on a significant oversimplification of how Scripture describes the work of each Person of the Trinity. As even Reformed scholar Curt Daniel has acknowledged, there are both general and particular aspects to the work of each Person. The Father loves all people as creatures (Matt. 5:45) but gives special saving love to the elect. The Spirit convicts and calls all people (John 16:8) but effectually calls only those who come to faith. Similarly, the Son died for all people but in a special manner for those who would believe. This is not Trinitarian discord — it is the same pattern of general and particular that we see in the work of the Father and the Spirit.21
In fact, it is limited atonement that creates a more serious problem for the unity of God's character. If Christ died only for the elect, then God is offering the gospel to people for whom no atonement has been made. When a preacher stands before a crowd and says, "Christ died for your sins — believe in Him and be saved," the limited atonement position requires that this offer is not genuinely available to everyone in the crowd. For the non-elect, there is no atonement to receive. How can a God of truth make a genuine offer of something that does not exist?22
Moreover, 2 Corinthians 5:18–20 teaches that God Himself is making the appeal through the preaching of the gospel: "God making his appeal through us. We implore you on behalf of Christ, be reconciled to God." God is the one making the offer. If He has limited the atonement to the elect alone, then His universal offer of salvation — addressed to all people without distinction — cannot be made in good faith. Only unlimited atonement preserves the sincerity of God's offer in the gospel.23
Some Calvinists argue that if Christ died for the sins of all people, then all people must be saved. If the price has been paid for everyone, how can anyone be lost? This would lead to universalism — the idea that all people will ultimately be saved. Since universalism is unbiblical, unlimited atonement must be false.
But this argument confuses the extent of the atonement with the application of the atonement. No one is automatically saved by Christ's death on the cross. The atonement provides the objective basis for salvation — but it must be received by faith. As we discussed above, propitiation accomplished is not the same as propitiation applied. Christ bore the sins of all, making salvation genuinely available to every person. But those who refuse to trust in Christ — who reject the offer of grace — remain under condemnation, not because the atonement was insufficient for them but because they have rejected the gift (John 3:18, 36; 2 Thess. 1:8).
Craig makes this point clearly in his treatment of the "problem of application." He argues that Christ's death accomplishes our potential redemption, which is then actualized progressively throughout history as people come to faith. "Like a pardon that has been issued and yet refused, Christ's redeeming work remains inefficacious in the lives of those who refuse him."24 Even Reformed thinkers recognize a distinction between redemption accomplished and redemption applied. The atonement is accomplished at the cross; it is applied when individuals believe. This distinction, as Craig observes, is "vital because otherwise the elect would be born redeemed" — which contradicts Scripture's teaching that believers were once "children of wrath like the rest of mankind" (Eph. 2:3).25
Unlimited atonement does not lead to universalism. It leads to what we might call universal availability: salvation is genuinely available to every person who has ever lived, because Christ died for every person. But availability is not the same as inevitability. The free response of faith is required for the atonement's benefits to be applied.
We touched on John 17 above, but the intercession argument deserves a bit more attention because it is commonly cited. The argument runs: Jesus intercedes only for the elect (John 17:9); since intercession and atonement go together, Christ must have died only for the elect.
This argument faces multiple problems. First, as we have seen, John 17:21 and 23 show that Jesus does pray for the world — that it might believe and know. So the premise that Jesus prays only for the elect is false. Second, even if we grant that Jesus' specific petitions in John 17:6–19 are directed to believers, it does not follow that the scope of His atonement is limited to them. Intercession and atonement are related but distinct. One can die for all while praying specific prayers for specific people. A soldier might give his life to protect his entire platoon while writing a letter home specifically to his wife. The letter's particular audience does not limit the scope of his sacrifice.26
Third, the argument commits what we might call the "collapsing fallacy" — it collapses the distinct work of Christ's intercession into His work of expiation, as though they must have identical scopes. But Scripture does not require this. Christ's atoning death is for all (1 John 2:2); His ongoing heavenly intercession is for those who draw near to God through Him (Heb. 7:25). These are complementary truths, not contradictory ones.
The Logic of the Gospel Offer: If limited atonement is true, God is offering salvation through the gospel to people for whom no atonement exists. This undermines the sincerity of the gospel offer. Only unlimited atonement can ground the genuine, universal proclamation: "Christ died for your sins — believe and be saved."
Many Calvinists who hold to limited atonement will say that Christ's death is "sufficient for all, efficient for the elect." This formula, which originated in the medieval period, sounds like a reasonable middle position. But Allen has shown that when limited atonement proponents use the word "sufficient," they mean something very different from what it normally means.27
On the limited atonement view, "sufficient for all" means only that Christ's death has infinite intrinsic value — it could have covered the sins of the whole world if God had intended it to. But since God did not intend it for the non-elect, no actual atonement exists for them. This is not genuine sufficiency. It is hypothetical sufficiency at best.
Think about what this means. If limited atonement is true, then the non-elect are in exactly the same position as if Christ had never died at all — there is no atonement for their sins. They could not be saved even if they wanted to be. Allen lays out the devastating implications with characteristic clarity:
If limited atonement is correct, Jesus did not substitute Himself on the cross for the sins of the non-elect. If this is the case... it is impossible that the non-elect could ever be saved since there is no atonement made for their sins. They are in the same unsavable state they would be in if Jesus had never come at all.28
Allen's point is sharp and I believe unanswerable. An atonement that does not actually cover someone's sins cannot be called "sufficient" for that person in any meaningful sense. You cannot offer someone something that does not exist for them. True sufficiency means that Christ's death actually provides an atonement for the sins of every person, such that if any person believes, they will find a real, existing atonement available for them — not a hypothetical one.
Allen further presses the question: suppose one of the non-elect were to believe — could they be saved? On the limited atonement position, the answer is no, because no satisfaction for their sins exists. But Scripture teaches that "whoever believes in him should not perish but have eternal life" (John 3:16). The "whoever" (pas ho pisteuōn, πᾶς ὁ πιστεύων) is genuinely universal. It means anyone, anywhere, without exception. This only makes sense if there is a real atonement available for anyone who believes.29
One of the strongest arguments against limited atonement is historical. For the first fifteen centuries of church history, the universal scope of Christ's death was essentially uncontested. The question of "limited versus unlimited atonement" did not arise as a serious debate until the post-Reformation period. Allen notes that prior to the Reformation era, "in the entire history of the church, there is evidence of only three people who seriously questioned that Christ died for the sins of all people."30
The Church Fathers — both Eastern and Western — consistently taught that Christ died for all people. As we demonstrated in Chapters 13–15 of this book, the patristic witness is overwhelmingly in favor of a universal scope. The Fathers spoke freely of Christ's death "for all," "for the whole world," and "for every human being." The idea that Christ died only for a pre-selected group of elect individuals is entirely absent from patristic theology.
Even the first generation of Protestant Reformers affirmed unlimited atonement. Allen argues that "Luther, Calvin, and Zwingli, along with all their colleagues and followers, held to unlimited atonement."31 This may come as a surprise, especially regarding Calvin, who is often assumed to have taught limited atonement. But careful scholarship has shown that Calvin's own statements about the extent of the atonement are consistently universal. He taught that Christ died for all people and that the gospel offer is genuinely universal. Limited atonement as a clearly articulated doctrine did not appear until Theodore Beza and William Perkins in the late sixteenth century — a generation after Calvin.32
This historical point is significant. Limited atonement is not the historic position of the church. It is a relatively late theological development, arising from a particular set of logical deductions about the relationship between election, atonement, and efficacy. The overwhelming weight of the Christian tradition — from the Apostolic Fathers through the medieval period through the earliest Reformers — supports unlimited atonement.
Even within the Reformed tradition itself, limited atonement has always been the most contested of the five points. The Amyraldian or "four-point Calvinist" position — named after the French Reformed theologian Moïse Amyraut (1596–1664) — affirms total depravity, unconditional election, irresistible grace, and perseverance of the saints while rejecting limited atonement. Amyraut argued that Christ died for all people, but that God's efficacious grace is applied only to the elect. This position was held by numerous theologians within the Reformed tradition and was never officially condemned by the Reformed churches as heretical.
Richard Baxter, the great Puritan pastor; John Davenant, the English delegate to the Synod of Dort; James Ussher, the Archbishop of Armagh; and many others within the Reformed tradition have affirmed unlimited atonement while holding to the other points of Calvinism. Even at the Synod of Dort (1618–1619) — the very council that codified the five points — the delegates did not unanimously endorse limited atonement in the strict sense. The Canons of Dort affirm that Christ's death is "of infinite worth and value, abundantly sufficient to expiate the sins of the whole world" (Second Head, Article 3). While the Canons go on to speak of God's "purpose and intention" to save the elect specifically (Article 8), the language about sufficiency was deliberately kept broad enough to accommodate moderate positions.33
The fact that so many within the Reformed tradition itself have rejected limited atonement suggests that it is not an essential corollary of Reformed theology. One can affirm God's sovereignty in salvation, unconditional election, the effectual call, and the perseverance of the saints — and still believe that Christ died for all people.
Historical Fact: The universal scope of the atonement was virtually uncontested for the first fifteen centuries of church history. Even the first generation of Reformers — including Luther and Calvin — affirmed that Christ died for all people. Limited atonement is a relatively late development, and even within the Reformed tradition, it has always been the most contested of the five points.
There is one more argument against limited atonement that I believe deserves careful attention, and it concerns the character of God Himself. As we argued in Chapter 3, the love of God is not merely something God does — it is something God is (1 John 4:8). God's nature is such that He loves all people and genuinely desires their salvation (1 Tim. 2:4; 2 Pet. 3:9; Ezek. 33:11). The cross is the supreme expression of that love (Rom. 5:8; John 3:16).
Limited atonement cuts across this biblical picture. If God determined that Christ would die only for the sins of the elect, then He loves the elect in a fundamentally different way than He loves the non-elect. For the elect, He provided an actual atonement. For the non-elect, He provided nothing — no real means of salvation at all. They are, as Allen puts it, "in the same unsavable state they would be in if Jesus had never come at all."34
How can we say, as Scripture does, that "God so loved the world" (John 3:16) if God never made an atonement available for most of the people in the world? How can we preach "God loves you and Christ died for you" to anyone, if we know that for many of our hearers, this is simply not true? The love of God, as revealed in the cross, demands a universal scope. God is not a God who plays favorites with His offer of redemption. He genuinely loves every human being He has made, and He has provided a genuine atonement for every one of them.
This does not mean that God's love operates in exactly the same way toward all people. Clearly, God has a special relationship with those who come to Him through faith. There is a "saving love" that believers experience — adoption, justification, indwelling by the Spirit — that unbelievers do not yet share. But this special love is not a denial of the universal love. God provides the atonement for all; He offers the gospel to all; He genuinely desires the salvation of all. Those who receive His grace by faith enter into the special blessings of covenant relationship. Those who reject it are condemned, not because God did not love them or did not provide for them, but because they refused the gift.
We have touched on this briefly, but the implications for gospel proclamation are so significant that they deserve fuller treatment. If limited atonement is true, then preachers and missionaries face a serious integrity problem.
Consider what happens when a preacher stands before a crowd of unbelievers and proclaims the gospel. The preacher says: "Christ died for sinners. He died for you. If you will believe in Him, your sins will be forgiven." On the unlimited atonement view, this message is straightforward and honest. Christ really did die for every person in that room, and if any of them believe, there is a genuine atonement available for them.
But on the limited atonement view, the preacher cannot honestly say "Christ died for you" to the crowd, because for many of them — the non-elect — Christ did not die for them. The preacher does not know who is elect and who is not, so the usual response from limited atonement proponents is: "We preach the gospel to all because we don't know who the elect are." But this misses the deeper problem. The issue is not our ignorance of who the elect are. The issue is that we are offering something to all people that does not exist for all people. This is not a genuine offer — it is a charade.35
Second Corinthians 5:18–20 makes clear that it is God Himself who is making the offer of salvation through the church. Paul writes: "God was reconciling the world to himself in Christ, not counting their trespasses against them, and entrusting to us the message of reconciliation. Therefore, we are ambassadors for Christ, God making his appeal through us." If God has limited the atonement to the elect alone, then His own appeal to the world — "Be reconciled to God" — is not made in good faith to the non-elect, because there is no reconciliation available for them.36
Allen raises the further question: "If Christ did not die for the sins of all people, what exactly are unbelievers guilty of rejecting?" Unbelief, by definition, is the rejection of something that has been offered. If no atonement exists for the non-elect, there is nothing for them to reject. They cannot be guilty of refusing a gift that was never given.37
Only unlimited atonement preserves the sincerity and genuineness of the gospel offer. Because Christ died for all, the gospel can be offered to all in good faith. "Whoever believes" (John 3:16) means exactly what it says: anyone. Not anyone who happens to be elect. Anyone. Period.
I want to take a moment to engage with a particularly careful defense of limited atonement found in Jeffery, Ovey, and Sach's Pierced for Our Transgressions. This book is one of the finest evangelical defenses of penal substitutionary atonement ever written, and I have drawn on it extensively throughout this study. On penal substitution, we are in strong agreement. On the extent of the atonement, we respectfully disagree.
Jeffery, Ovey, and Sach argue that particular redemption best preserves the efficacy of Christ's death. They worry that universal atonement "pictures Christ's atonement as a ship with all humanity on board that drops anchor a mile or two out to sea, leaving its passengers to swim ashore," while particular redemption "depicts it as a lifeboat that unfailingly conveys the whole company of God's elect... all the way from the storms of death and condemnation to the shelter of God's eternal kingdom."38 They also argue that unlimited atonement "divides the persons of the Trinity" because the Son's intention to save all would be in conflict with the Father's election of only some.39
These are thoughtful arguments, but I have already shown why I believe they do not hold up. The "ship" analogy, while vivid, presents a false dichotomy. Unlimited atonement does not say the cross merely "makes salvation possible" without actually accomplishing anything. The cross accomplishes the objective reconciliation of God and humanity (2 Cor. 5:19). It removes the legal barriers to salvation. It provides a real, complete, and sufficient atonement for every human being. What it does not do is force that atonement on anyone. Faith is the means by which the accomplished atonement is applied. This is not a weakening of the cross — it is a recognition that God respects human freedom and genuinely invites a response.
As for the Trinitarian argument, I have already noted that there are general and particular dimensions to the work of each Person. The Father's common grace, the Spirit's universal conviction, and the Son's universal atonement are all general aspects of God's saving work. The Father's election, the Spirit's effectual call, and the Son's special love for the church are the particular aspects. These complement each other rather than contradicting each other.
I also want to note that Jeffery, Ovey, and Sach themselves acknowledge, as Packer did, that their defense of particular redemption is offered as a response to Eleonore Stump's philosophical challenge about the coherence of penal substitution with universal atonement.40 But as Craig has argued persuasively, Stump's argument is really aimed at limited atonement, not at penal substitution itself. The solution to Stump's challenge is not to limit the atonement but to distinguish between redemption accomplished and redemption applied — which even Reformed theologians already do.41
Redemption Accomplished vs. Applied: Even Reformed theologians distinguish between what Christ accomplished at the cross (objective atonement) and how it is applied to individuals (through faith). This distinction — not limited atonement — is the proper answer to the question of why all are not saved despite Christ dying for all.
Let me draw the threads together and state my own position clearly.
I believe that Christ died for all people without exception. His death provides a genuine, actual, and sufficient atonement for every human being who has ever lived or ever will live. The scope of the atonement is universal. This is the clear and consistent teaching of Scripture (as argued in Chapter 30), and it has been the majority position of the church throughout history.
At the same time, the application of the atonement is particular. Not everyone will be saved, because not everyone will respond to the gospel in faith. The atonement is objectively accomplished for all but subjectively appropriated through faith (as discussed in Chapter 29). Those who believe receive the benefits of Christ's death — justification, reconciliation, redemption, adoption. Those who refuse the gift remain under condemnation, not because the atonement was insufficient for them but because they have rejected the grace that was genuinely offered to them.
This position preserves everything that limited atonement seeks to protect — the efficacy of the cross, the sovereignty of God, the certainty of salvation for believers — while also preserving the universal love of God, the genuineness of the gospel offer, and the clear testimony of Scripture that Christ died "for all" and "for the whole world."
I also want to note briefly how this connects to my broader theological framework. As I have mentioned at various points in this book, I believe in the possibility of a postmortem opportunity for salvation — that those who never had a genuine chance to hear and respond to the gospel in this life may receive that opportunity after death, with the last chance to receive Christ being at or during the last judgment. The universal scope of the atonement is essential to this hope. If Christ died only for the elect, then there is no basis for offering salvation to those who die without having heard the gospel. But if Christ died for all, then His atoning work covers every human being who has ever lived — including those in cultures and ages where the gospel was never preached. The atonement's universal scope provides the objective foundation for the hope that God's grace may reach beyond the boundaries we see in this present age.
One concern that I want to address directly is whether rejecting limited atonement somehow weakens the case for penal substitutionary atonement. Some defenders of particular redemption, like Jeffery, Ovey, and Sach, have suggested that PSA and limited atonement fit together naturally, and that unlimited atonement creates tensions for penal substitution.
I respectfully but firmly disagree. Penal substitutionary atonement does not require limited atonement. Christ can bear the penalty for the sins of all people without that penalty being automatically applied to all people. The atonement provides the basis for forgiveness; faith provides the condition for its reception. There is no logical contradiction in saying, "Christ bore the judicial consequences of every person's sin, and this atoning work is applied to those who trust in Him."
Craig makes this point clearly. The objection that Christ's suffering for all sins would imply universal salvation "need not be met by affirming the doctrine of limited atonement. For Christ's atoning death may be taken to have potentially accomplished the redemption of all men, but this redemption is actualized progressively throughout history, only as people come to be united with Christ through repentance and faith."42
In fact, I would argue that unlimited atonement strengthens the case for penal substitution, because it makes the gospel offer genuinely universal and grounded in an actual atonement for every person. When we tell someone, "Christ died for your sins," we are not speaking hypothetically. We are speaking of an accomplished reality. That reality awaits their response of faith — but it is real, not theoretical.
Throughout this book, I have argued that penal substitutionary atonement, rightly understood within a Trinitarian framework of divine love, is the central facet of the atonement. I have also argued that the atonement has a universal scope — Christ died for all. These two convictions are not in tension. They are complementary truths that together reveal the full glory of the cross: God Himself, in the Person of His Son, bore the judicial consequences of the sins of every human being, so that anyone who trusts in Him may be forgiven and reconciled to God.
We have covered a great deal of ground in this chapter. Let me summarize our findings.
The Calvinist doctrine of limited atonement rests primarily on logical deductions rather than clear exegetical evidence. Not a single text of Scripture says that Christ died only for the elect. The "particular" texts (John 10:11; Eph. 5:25; Matt. 1:21) express Christ's special love for His own but do not exclude others — to read them as exclusive is to commit the negative inference fallacy. Meanwhile, the "universal" texts (1 John 2:2; 1 Tim. 2:6; Heb. 2:9; John 3:16; 2 Pet. 3:9) are most naturally read as genuinely universal.
The five major arguments for limited atonement — the double payment argument, Owen's trilemma, the Trinitarian unity argument, the universalism entailed argument, and the intercession argument — all fail upon careful examination. The double payment argument rests on a commercial model of the atonement that Scripture does not support. Owen's trilemma proves too much on its own terms. The Trinitarian unity argument oversimplifies the general and particular aspects of the work of each divine Person. The universalism entailed argument confuses the extent of the atonement with its application. And the intercession argument misreads John 17, which actually shows Jesus praying for the world's salvation.
Historically, unlimited atonement has been the majority position of the church from the patristic era through the Reformation and beyond. Even within the Reformed tradition, limited atonement has always been the most contested of the five points, with many faithful Reformed thinkers rejecting it.
The practical implications of limited atonement are also deeply troubling. It undermines the genuineness of the gospel offer, makes it impossible for preachers to tell every person honestly that "Christ died for your sins," and portrays a God who offers salvation to people for whom no salvation exists. None of this is consistent with the biblical picture of a God who "so loved the world" and "desires all people to be saved."
I want to close on a pastoral note. The doctrine of limited atonement, when pressed to its logical conclusions, has a chilling effect on evangelism, assurance, and the comfort of the gospel. If I cannot tell every person I meet that Christ died for them, then the good news is not quite as good as it appears. But the Scriptures give us a bigger, more generous, more beautiful gospel than limited atonement allows. Christ died for all. The offer is genuine. The atonement is real. And "whoever believes in him should not perish but have eternal life" (John 3:16).
That is the gospel I find in Scripture, and that is the gospel I commend to you.
1 David L. Allen, The Atonement: A Biblical, Theological, and Historical Study of the Cross of Christ (Nashville: B&H Academic, 2019), 155. Allen notes that limited atonement has always been the most controversial of the five points of Calvinism. ↩
2 Steve Jeffery, Michael Ovey, and Andrew Sach, Pierced for Our Transgressions: Rediscovering the Glory of Penal Substitution (Wheaton, IL: Crossway, 2007), 270–71. They cite Roger Nicole's clarification that the question is about the design and intention of the atonement, not its intrinsic value. ↩
3 J. I. Packer, "What Did the Cross Achieve? The Logic of Penal Substitution," Tyndale Bulletin 25 (1974): 3–45. Jeffery, Ovey, and Sach cite Packer's argument at length: Jeffery, Ovey, and Sach, Pierced for Our Transgressions, 273–74. ↩
4 Allen, The Atonement, 156. Allen writes: "There is no single text of Scripture asserting Jesus died only for the sins of the elect. Limited atonement is a doctrine in search of a text." ↩
5 Allen, The Atonement, 157–58. Allen demonstrates how the "for the sheep" / "for the church" texts commit the negative inference fallacy when used to support limited atonement. ↩
6 Allen, The Atonement, 172–74. Allen draws on the work of David Ponter to show that "world" in John 17:21, 23 refers to the unbelieving world, and the verbs "believe" and "know" carry their normal salvific meaning. ↩
7 Allen, The Atonement, 172. Allen summarizes Dekker's argument that Jesus' specific petitions in John 17:6–19 concern the needs of believers, but His broader concern in the chapter includes the world's salvation. ↩
8 Allen, The Atonement, 175. Allen notes that the argument can be inverted: if Christ prays for a person, He must have died for that person. John 17:21 and 23 show Christ praying for the world. ↩
9 Allen, The Atonement, 158–62. See also the extensive treatment of 1 John 2:2 and the meaning of "the whole world" in Chapter 30 of this book. ↩
10 John Owen, The Death of Death in the Death of Christ (Edinburgh: Banner of Truth, 1959), 173–74. Cited in Jeffery, Ovey, and Sach, Pierced for Our Transgressions, 269. ↩
11 Allen, The Atonement, 163–65. Allen provides the restaurant analogy and the distinction between commercial and criminal debt. ↩
12 Allen, The Atonement, 164. ↩
13 William Lane Craig, Atonement and the Death of Christ: An Exegetical, Historical, and Philosophical Exploration (Waco, TX: Baylor University Press, 2020), chap. 10, "Penal Substitution: Its Justification," under discussion of whether Christ's suffering implies universal salvation. Craig argues that Christ's death potentially accomplished redemption for all, which is actualized through faith. ↩
14 Charles Hodge, Systematic Theology (Grand Rapids: Eerdmans, 1952), 2:471. Cited in Allen, The Atonement, 165. ↩
15 Hodge, Systematic Theology, 2:545. Cited in Allen, The Atonement, 169–70. ↩
16 Allen, The Atonement, 160–62. Allen's argument about the noun-verb distinction in hilasmos and paraklētos is one of the most insightful technical contributions to this debate. ↩
17 Owen, The Death of Death, 61–62. See Allen, The Atonement, 166–69 for the full critique of Owen's trilemma. ↩
18 Allen, The Atonement, 167–68. Allen draws heavily on the work of Neil Chambers in demonstrating that Owen's trilemma proves too much. ↩
19 Allen, The Atonement, 167–68. Chambers argues that Owen engaged in "polemical reductionism" by failing to distinguish between unbelief as a sin and unbelief as a state. ↩
20 Allen, The Atonement, 169. Allen argues that Owen's trilemma "necessarily operates on the assumption that there was a quantitative imputation of sins to Christ," which is not how imputation functions biblically. ↩
21 Allen, The Atonement, 170. Allen cites Curt Daniel's observation that there are general and particular aspects of the work of each Person of the Trinity, and that a strictly limited atonement would logically lead to hyper-Calvinism. ↩
22 Allen, The Atonement, 179–80. Allen argues that limited atonement undermines the genuineness of the gospel offer. ↩
23 Allen, The Atonement, 180. Allen notes that 2 Corinthians 5:18–20 shows God Himself making the offer through the church, which requires that the atonement is genuine for all to whom the offer is made. ↩
24 Craig, Atonement and the Death of Christ, chap. 10, "Penal Substitution: Its Justification," concluding remarks on the scope of the atonement. ↩
25 Craig, Atonement and the Death of Christ, chap. 10, "Penal Substitution: Its Justification." Craig argues that the accomplished/applied distinction is vital because "otherwise the elect would be born redeemed." ↩
26 Allen, The Atonement, 171–72. Allen argues that collapsing Christ's intercession into His expiation merely begs the question. ↩
27 Allen, The Atonement, 177–79. Allen distinguishes between "infinite/universal/extrinsic sufficiency" (which means the atonement is actually available for all) and "limited/intrinsic sufficiency" (which means the atonement could hypothetically cover all but actually does not). ↩
28 Allen, The Atonement, 179. ↩
29 Allen, The Atonement, 180–81. ↩
30 Allen, The Atonement, 155. ↩
31 Allen, The Atonement, 155. Allen states that the first-generation Reformers on the continent and in England all held to unlimited atonement. ↩
32 Allen, The Atonement, 155. Allen dates the clear advocacy of limited atonement to Theodore Beza and William Perkins in the late sixteenth century. See also David L. Allen, The Extent of the Atonement: A Historical and Critical Review (Nashville: B&H Academic, 2016) for the comprehensive treatment of this historical question. ↩
33 The Canons of Dort, Second Head of Doctrine, Articles 3 and 8. See also W. Robert Godfrey, "Reformed Thought on the Extent of the Atonement to 1618," Westminster Theological Journal 37 (1975): 133–71. ↩
34 Allen, The Atonement, 179. ↩
35 Allen, The Atonement, 179–82. Allen's extended argument about the implications of limited atonement for preaching and the gospel offer is among the most practically significant sections of his book. ↩
36 Allen, The Atonement, 180. ↩
37 Allen, The Atonement, 180. Allen asks: "If Christ did not die for the sins of all people, what exactly are unbelievers guilty of rejecting?" ↩
38 Jeffery, Ovey, and Sach, Pierced for Our Transgressions, 274. ↩
39 Jeffery, Ovey, and Sach, Pierced for Our Transgressions, 273. ↩
40 Jeffery, Ovey, and Sach, Pierced for Our Transgressions, 269–70. Their defense of particular redemption is explicitly framed as a response to the philosophical challenge raised by Eleonore Stump. ↩
41 Craig, Atonement and the Death of Christ, chap. 10, "Penal Substitution: Its Justification." Craig argues that Stump's objection is really against limited atonement, not penal substitution, and notes that the accomplished/applied distinction resolves the issue without requiring limited atonement. ↩
42 Craig, Atonement and the Death of Christ, chap. 15, "Concluding Remarks." Craig writes that Christ's atoning death may be taken to have potentially accomplished the redemption of all, actualized as people come to faith. ↩
43 John R. W. Stott, The Cross of Christ (Downers Grove, IL: InterVarsity Press, 2006), 149. Stott speaks of the "self-substitution of God" in a way that is fully compatible with universal atonement — God Himself, in the Person of His Son, bears the penalty of human sin for all. ↩
44 Leon Morris, The Apostolic Preaching of the Cross, 3rd ed. (Grand Rapids: Eerdmans, 1965), 60–64. Morris's work on the propitiatory nature of Christ's death supports the penal dimension of the atonement without requiring limited atonement. ↩
45 I. Howard Marshall, Aspects of the Atonement: Cross and Resurrection in the Reconciling of God and Humanity (Milton Keynes: Paternoster, 2007), 60–82. Marshall defends unlimited atonement as the clear teaching of the New Testament while affirming the substitutionary and penal dimensions of the cross. ↩
46 Wayne Grudem, Systematic Theology: An Introduction to Biblical Doctrine (Grand Rapids: Zondervan, 1994), 594–603. Grudem presents the case for particular redemption but acknowledges the strength of the arguments for unlimited atonement. ↩
47 D. A. Carson, The Difficult Doctrine of the Love of God (Wheaton, IL: Crossway, 2000), 73–79. Carson acknowledges that the Greek word kosmos never means "all the elect" collectively in the New Testament, at least in the Johannine corpus. Cited in Allen, The Atonement, 159. ↩
48 Robert Letham, The Work of Christ (Downers Grove, IL: InterVarsity Press, 1993), 228–41. Letham provides a balanced discussion of the extent of the atonement from within the Reformed tradition. ↩
49 Henri Blocher, "The Scope of Redemption and Modern Theology," Scottish Bulletin of Evangelical Theology 9 (1991): 80–103. Blocher argues for a carefully nuanced position on the extent of the atonement. ↩
50 Thomas R. Schreiner, "Penal Substitution View," in The Nature of the Atonement: Four Views, ed. James Beilby and Paul R. Eddy (Downers Grove, IL: IVP Academic, 2006), 67–98. Schreiner defends PSA without requiring limited atonement, acknowledging that the question of extent is distinct from the question of nature. ↩
Allen, David L. The Atonement: A Biblical, Theological, and Historical Study of the Cross of Christ. Nashville: B&H Academic, 2019.
Allen, David L. The Extent of the Atonement: A Historical and Critical Review. Nashville: B&H Academic, 2016.
Blocher, Henri. "The Scope of Redemption and Modern Theology." Scottish Bulletin of Evangelical Theology 9 (1991): 80–103.
Carson, D. A. The Difficult Doctrine of the Love of God. Wheaton, IL: Crossway, 2000.
Craig, William Lane. Atonement and the Death of Christ: An Exegetical, Historical, and Philosophical Exploration. Waco, TX: Baylor University Press, 2020.
Godfrey, W. Robert. "Reformed Thought on the Extent of the Atonement to 1618." Westminster Theological Journal 37 (1975): 133–71.
Grudem, Wayne. Systematic Theology: An Introduction to Biblical Doctrine. Grand Rapids: Zondervan, 1994.
Hodge, Charles. Systematic Theology. 3 vols. Grand Rapids: Eerdmans, 1952.
Jeffery, Steve, Michael Ovey, and Andrew Sach. Pierced for Our Transgressions: Rediscovering the Glory of Penal Substitution. Wheaton, IL: Crossway, 2007.
Letham, Robert. The Work of Christ. Downers Grove, IL: InterVarsity Press, 1993.
Marshall, I. Howard. Aspects of the Atonement: Cross and Resurrection in the Reconciling of God and Humanity. Milton Keynes: Paternoster, 2007.
Morris, Leon. The Apostolic Preaching of the Cross. 3rd ed. Grand Rapids: Eerdmans, 1965.
Owen, John. The Death of Death in the Death of Christ. Edinburgh: Banner of Truth, 1959.
Packer, J. I. "What Did the Cross Achieve? The Logic of Penal Substitution." Tyndale Bulletin 25 (1974): 3–45.
Schreiner, Thomas R. "Penal Substitution View." In The Nature of the Atonement: Four Views, edited by James Beilby and Paul R. Eddy, 67–98. Downers Grove, IL: IVP Academic, 2006.
Stott, John R. W. The Cross of Christ. Downers Grove, IL: InterVarsity Press, 2006.