When Was Revelation Written?

Chapter 4 and Appendix 1 of Dismantling Religion

Everything in Revelation turns on one question of dating. If John wrote in 95 AD, the book looks past the destruction of Jerusalem and forward to a future still pending. If he wrote before the city fell, the same visions describe a war his first readers could see approaching — and the New Testament’s timetable stops being a two-thousand-year delay.

This is not a technicality for scholars. It decides whether the book is chiefly about something that already happened or something that has not. So it is worth arguing properly rather than asserting, and the case divides into what the earliest Christians said and what the text itself assumes. Both point the same way.

The case for 95 AD, and why it is weaker than it looks

The late date rests almost entirely on one sentence from Irenaeus, writing in the late second century: that the apocalyptic vision was seen “towards the end of Domitian’s reign.” Domitian ruled from 81 to 96 AD. From that sentence, a whole chronology is built.

But the sentence is doing more work than it can bear.

First, the surviving text is a poor paraphrase rather than a quotation, and the Greek permits a different subject: the phrase may mean that John himself was seen towards the end of Domitian’s reign, not that the vision was given then. A man can be seen long after he wrote.

Second, Irenaeus is demonstrably careless elsewhere. He affirms that Jesus lived to around fifty years of age — a claim no historian today accepts and which contradicts the gospels he was defending. A writer who misremembers the length of Christ’s life is not a safe anchor for a single date.

Third, the word itself may have been misread. Robert Young argued in 1885 that the Greek term behind “Domitian” could originally have meant Domitius — that is, Nero, whose full name was Nero Claudius Caesar Drusus Germanicus, born Lucius Domitius Ahenobarbus. A later copyist, seeing an unfamiliar form, substituted the emperor he knew. And an early Syriac title of Revelation places John’s banishment under Nero, around 68 AD — not under Domitian at all.

So the late date hangs on a paraphrase, from a writer with a track record of error, of a word that may have been changed in transmission. That is not nothing. But it is not the kind of evidence you would build a theology on.

What the early church actually said

Once the Irenaeus sentence stops being decisive, the other witnesses come into view, and they are remarkably consistent.

The Muratorian Canon, written around 170 AD and the earliest surviving list of New Testament books, places John’s writing before Paul’s letters were complete — and Paul died around 67–68 AD. It states explicitly that Revelation was written during Nero’s reign.

The Monarchian Prologues, from somewhere between 250 and 350 AD, say the same thing.

The Syriac New Testament, in a tradition reaching back before the sixth century, and the Aramaic Peshitta of the fifth, both place the book before 70 AD.

Tertullian links John’s banishment to the same persecution that killed Peter and Paul — which is to say, under Nero, between 64 and 68 AD.

Jerome preserves a detail that is difficult to invent: that Nero first had John plunged into boiling oil, and, when that failed, exiled him to Patmos.

Andreas of Caesarea, writing around 500 AD, applies Revelation 6:12–13 — the sun blackened, the stars falling — to the destruction of Jerusalem.

Arethas, archbishop of Caesarea in the ninth and tenth centuries, says plainly that John wrote before 70 AD and that the prophecies concerned the Jewish–Roman war.

There is one witness that needs more care, because it is the one usually cited for the other side. Clement of Alexandria records that John left Patmos “after the death of the tyrant” to appoint bishops in the surrounding cities. Which tyrant?

The word was used of both Nero and Domitian, so on its own it settles nothing. But two things tilt it. First, Nero was the archetypal tyrant in early Christian memory — Apollonius of Tyana, no Christian, described Nero as more beastly even than Domitian. When early writers said “the tyrant” without qualification, the name that came to mind was Nero’s. Second, Clement’s narrative contains a physical detail that only makes sense on the earlier dating: he describes John, after his return, chasing a runaway convert on horseback through the countryside. That is an old man still riding, not a man of ninety. If John was in his early sixties under Nero, the detail fits; under Domitian he would have been near a hundred.

What the text itself assumes

External testimony is persuasive, but it is testimony — people can be wrong, and traditions drift. The harder evidence is internal, because it cannot be edited after the fact.

Nobody mentions that the temple has fallen

This is the point that ought to stop the argument. If Revelation was written in 95 AD, the temple in Jerusalem had been rubble for twenty-five years. The city had been levelled, the priesthood destroyed, the sacrificial system ended, and the whole shape of Jewish worship remade by catastrophe.

And yet John writes as though the temple were still there. Revelation 11 opens with a measuring rod and an instruction to measure the temple, the altar, and those who worship in it. He is not describing a memory. He is describing a building.

Consider what this book is about: judgment on a great city, the fall of Babylon, the vindication of the martyrs. If that city were Jerusalem, and Jerusalem had already fallen, the silence about its destruction would be inexplicable. You do not write four chapters about the doom of a city and never mention that it is already in ruins.

The usual answer is that Revelation is visionary rather than topographical — that John deals in symbols, so a literal temple need not be standing. But that proves too much. If the imagery is untethered from history, then none of it can be dated, including the judgments the book insists are near. John himself opens by saying these things must shortly come to pass, and closes by saying the time is at hand. A symbolic reading that removes the temple from history removes the urgency with it.

Forty-two months is not a round number

Revelation 11:2–3 gives a precise figure: the holy city will be trampled by the nations for forty-two months — three and a half years.

The siege of Jerusalem lasted, from the Roman campaign in Galilee in the spring of 67 AD to the fall of the city in September of 70 AD, almost exactly three and a half years. Luke 21:24 describes the same period in the same terms: Jerusalem surrounded by armies, Gentiles trampling the city, times fulfilled.

Symbolic numbers in apocalyptic literature tend to be symbolic precisely because they are round — a thousand years, seven seals, twelve tribes. Forty-two months is not round in that way. It is the kind of number you get from counting.

The great city is named, and it is not Rome

Revelation 17 describes a woman seated on many waters, drunk with the blood of the saints, and calls her Babylon the great. The natural assumption is Rome. But the text identifies her himself.

Revelation 11:8 speaks of the great city as the place “where their Lord was crucified.” Rome did not crucify Jesus. Jerusalem did.

The same verse calls the city “Sodom” — and Isaiah 1 uses precisely that name for Jerusalem, condemning a city whose worship had become a performance. Revelation 17:6 says the woman is drunk with the blood of the prophets, and Jesus had already charged Jerusalem with the blood of every prophet from Abel onward in Matthew 23:29–38, in the same discourse where he told that generation it would not pass away before these things came to pass.

So the harlot is not a cipher for imperial Rome in general. She is the city whose leaders handed over the Lord to be crucified, dressed in the borrowed splendour of a priesthood that had gone wrong — which brings us to the detail most readers skip.

The harlot is dressed like a high priest

Revelation 17:4 describes her clothing: purple and scarlet, gold and precious stones and pearls. Compare Exodus 28, where the high priest’s ephod is made of gold, blue, purple and scarlet, and the breastpiece is set with precious stones. The colours are not generic luxury. They are the colours of the sanctuary.

She holds a golden cup — and Josephus records that the temple vessels were of gold, including the cups used in its service (Wars 5.4.4).

And on her forehead she has a name written, blasphemous. Exodus 28:36 puts a gold plate on the high priest’s forehead inscribed “Holy to the Lord.” John gives the same placement to a woman whose inscription is blasphemy instead.

That is not how you describe a pagan city. That is how you describe a desecrated priesthood in the city that was meant to bear God’s name. The imagery only works if Jerusalem is the subject.

The vocabulary of desolation

When Revelation describes Babylon’s fall, it uses a word already loaded in Jewish ears: desolate.

Jesus used it of Jerusalem — your house is left to you desolate (Matthew 23:38), and Luke 21:20 has him describing Jerusalem surrounded by armies. Daniel used it of the abomination that makes desolate (Daniel 9:27). The word travelled, and it always meant one thing: covenant judgment, the sanctuary abandoned, the city given over to ruin.

John is not reaching for an evocative image. He is joining a chain of prophecy, and the last link in that chain had an address.

666, 616, and a name in two languages

Revelation 13 gives the number of the beast as 666. Some early manuscripts read 616 instead — a discrepancy that sceptics have used against the book and that turns out to be evidence for it.

Both numbers work on one name. Written in Hebrew letters, “Neron Caesar” totals 666. In Latin spelling, “Nero Caesar” totals 616. The variant was circulating in the second and third centuries, appearing in early Greek copies and Old Latin texts long before the Vulgate.

Why does that matter? Because scribes preserved both readings rather than correcting one to the other. They were not confused about who the beast was. They knew it was Nero, and they adjusted the arithmetic to the alphabet they were writing in.

And Nero was not incidental to the churches of Asia. He was the first emperor to turn full-scale persecution on Christians, beginning in 64 AD — the persecution in which Peter and Paul died. A book that encodes his name in its central riddle, written to seven churches under that pressure, is a book written into that crisis, not one looking back on it a generation later.

What follows if the date is early

If John wrote before 70 AD, and if the great city is Jerusalem, and if the forty-two months are the siege, then Revelation is not a map of a future anticlimax. It is a prophecy of something that happened within the lifetime of people who could still be asked about it.

That conclusion has consequences the rest of the book follows: if the judgment has come, then the question is no longer “when will this happen?” but “what came after?” — and the answer is the thousand years of Revelation 20, which Dismantling Religion argues can be dated, from 96 to 1096 AD, ending at the First Crusade.

That is a larger claim than the dating question, and it deserves to be tested separately. But it begins here, with a date.