8 min read

The Seventy Weeks of Daniel, Side by Side


Daniel has been reading Jeremiah's seventy years of exile, and he is on his knees confessing when Gabriel interrupts with a much bigger number.

The seventy weeks of Daniel are seventy sevens — 490 years on the year-for-a-day reading nearly everyone brings to the passage — in three pieces: 7, then 62, then 1. Daniel 9:24–27 gives that structure, the events closing each stretch, and a starting gun: the going forth of the commandment to restore and to build Jerusalem. What it never says is which commandment. Four candidates sit in the Old Testament, the outer two more than ninety years apart, and four further choices get made before any date appears. Below, all four sit in one table with the arithmetic shown and the assumptions labelled.

The seventy weeks of Daniel, in three pieces

Daniel is studying Jeremiah's seventy years of desolation, a number Chronicles ties to sabbath rest: the land lay empty until the land had enjoyed her sabbaths (2 Chronicles 36:21). Sevens are the unit before the angel opens his mouth.

The Hebrew is shavu'im shiv'im, seventy sevens. Translations split between "weeks" and "sevens" for a plain reason — the word counts sevens of something and never says sevens of what. One chapter later Daniel mourns three sevens of days, with days attached (Daniel 10:2): the same masculine plural, the unit spelled out. Given the chance to name it he names it, and in chapter 9 he leaves it open.

The year-for-a-day step is supplied by the reader, not the text, leaning on two passages that perform the conversion inside their own narratives (Numbers 14:34, Ezekiel 4:6). What makes it near-universal here is local: 490 days is about sixteen months, and nothing in verse 24 — atonement, everlasting righteousness, a finished transgression — fits inside sixteen months.

That gives 49 years, then 434, then 7. Traditions divide first over whether verse 24's six decreed things closed at the cross or remain outstanding, and every dispute below sits on that division.

Four decrees, four different clocks

Four Persian-era orders about Jerusalem or its temple sit in Scripture, each with serious defenders, each starting the clock elsewhere.

DecreeDateAuthorises+483 solar years+483 × 360 days
Cyrus (Ezra 1:1–4)538 BCReturn from exile; rebuild the temple55 BC62 BC
Darius I (Ezra 6:1–12)c. 520 BCConfirms Cyrus; funds the temple37 BC44 BC
Artaxerxes I to Ezra (Ezra 7:11–26)458/457 BCTemple funds; magistrates and judgesAD 27AD 20
Artaxerxes I to Nehemiah (Neh 2:1–8)445/444 BCTimber for the city wall and gatesAD 39AD 32/33

Same text. Same arithmetic. The far ends of that table stand a century apart.

Cyrus and Darius fall away quickly: Ezra 1 concerns the house of God, not the city, and Darius merely confirms Cyrus. The two live options trade the same argument. Ezra 7 grants no wall, but it does grant courts, judges and a working legal order — arguably what restore means applied to a people rather than masonry. Nehemiah 2 is the mirror image: no law, but explicit royal authority over the city's walls and gates, the closest verbal match in Scripture to "restore and build Jerusalem." Historicists, Millerite and Adventist, take Ezra 7. Dispensationalists take Nehemiah 2. The double dates are not sloppiness — Artaxerxes I's regnal years run from spring or autumn, with or without his accession year, shifting his seventh and twentieth years by one.

That silent choice carries more weight than any sum built on it: William Miller counted 2,300 years from this same Ezra decree to 1844, and why end-times date predictions fail follows what it cost.

Where seven years go missing

A "prophetic year" of 360 days is shorter than a solar one — the most under-explained move in the debate. Run the multiplication: 483 × 360 = 173,880 days. Divide by 365.2422 days, the average length of a tropical year, and you get 476.07: a shade over 476 years rather than 483. The units disagree by very nearly seven years across that span. And an average is all 365.2422 is — real years differ, a calendar reform sits inside the interval, and the AD 32/33 column is close enough to size the gap, nowhere near sharp enough to name a Friday.

Which is why 445 BC yields AD 39 in one column, AD 32 in the other. Move the anchor a year and the answer moves a year. Change the unit and it moves seven.

The case for that unit gets its own hearing, with the day-for-a-year precedents, in the 360-day prophetic year. What matters here is narrower: the interpreter chooses it, Daniel 9 does not print it.

The accent mark that halves the count

Open two English Bibles at Daniel 9:25 and you may find two different prophecies. The King James: unto the Messiah the Prince shall be seven weeks, and threescore and two weeks — one continuous 69. The ESV: to the coming of an anointed one, a prince, there shall be seven weeks. And for sixty-two weeks it shall be built again.

Not one count. Two — and two anointed figures rather than one.

The split turns on the Masoretic accents, the scribal tradition's own punctuation, which mark a pause after seven weeks. Read that way, the first 49 years run from Jerusalem's fall to somewhere near Cyrus, whom Isaiah 45:1 calls his anointed. The counterweight is honest: those accents are an old reading tradition, not an autograph, fixed by Tiberian scribes long after the consonants, and Hebrew can coordinate them the King James way instead. Undisputed: verse 25 puts a definite article on neither noun, reading "an anointed one, a prince." A fact about grammar, not a verdict about who is meant — and the New Testament reads Daniel as speaking to its own moment (Matthew 24:15).

Is there a gap before the seventieth week?

Nobody disputes the events, only where the sentence boundaries fall. Take the last two verses clause by clause and watch three traditions handle the same words.

"After threescore and two weeks shall Messiah be cut off, but not for himself." Futurist and historicist readers see one event here, a crucifixion at or just past the 69-week terminus, differing only over the anchor. The preterist sees another man: Onias III, the high priest killed shortly before Antiochus IV set a pagan altar in the temple in 167 BC. That gains a second-century book explaining second-century trouble and loses the arithmetic — sixty-two weeks from any decree in the table lands well past Onias, so the figures must be schematic.

"And the people of the prince that shall come shall destroy the city and the sanctuary." The futurist's firmest ground, and everybody else's hardest work. Jerusalem fell in AD 70, decades past any 69-week terminus in the table, and verse 27's single week has not opened yet — so something already stands between the sixty-nine and the seventieth. Historicists answer that verse 26 is summary rather than sequence, naming what follows the sixty-two weeks without dating it inside them. Christian preterists make AD 70 the point of the oracle.

"And he shall confirm the covenant with many for one week." Now the pronoun does the work. Read "he" as the anointed one of the previous verse and the historicist scheme snaps into place: from 457 BC the sixty-nine weeks reach AD 27, the seventieth runs to AD 34, and in the midst of the week he shall cause the sacrifice and the oblation to cease is the crucifixion at the midpoint, veil torn. Elegant, and bought by reaching past the nearer antecedent — the prince who destroys the city.

So the argument narrows to a seam a few words wide, between verse 26's destruction and verse 27's covenant. Whether it is a hinge, a pause, or twenty centuries of open ground, arithmetic cannot decide, and Day 4 does not claim to.

Day 4 makes no prophetic calculations and never will. What it holds is the furniture under the argument — a calendar whose timeline travels back toward the beginning, and the Scroll of Generations, tracing the generations from Adam, both on the subscription side above a free tier of today's devotional, your sacred date and the next Sabbath. Good for the shape of the ground, deliberately useless for naming a year. That chronology is assembled in what year is it on the Hebrew calendar.

Common questions

What do Daniel's seventy weeks mean?

Seventy sevens — read as 490 years — decreed over Jerusalem and Daniel's people, divided 7 + 62 + 1. Inside them: six accomplishments ending in atonement and everlasting righteousness, an anointed one cut off, a city and sanctuary destroyed, a covenant broken at the last week's midpoint. The events are stated plainly, the dates are not.

Which decree starts the count — 457 BC, 445 BC, or another?

Daniel does not say. The candidates are Cyrus (538 BC), Darius I (c. 520 BC), Artaxerxes to Ezra (458/457 BC), and Artaxerxes to Nehemiah (445/444 BC). Nehemiah's is the closest verbal match, alone concerning the city walls. Ezra's fits "restore" if restoration means courts and law. No outside evidence forces the choice.

Is there a gap before the seventieth week?

The text never mentions one, yet it places the city's destruction after the sixty-two weeks without fitting it inside the seventieth. Futurist readings treat that space as an open interval running to now. Historicist and preterist readings run the count unbroken and pay elsewhere — in the anchor date, or in treating the numbers as approximate.

How do the different traditions land on different dates?

Five things get decided before the sum begins: what a week means, which decree starts it, how long a year is, whether the count runs unbroken, and where boundaries fall inside a year. Changing the decree moves the answer by up to a century. Swapping 483 solar years for 483 × 360 days moves it by seven.

The nineteen verses before the numbers

Verses 3 to 19 are the part almost every treatment skips, and they outrun the prophecy in length. Daniel is fasting in sackcloth and ashes, confessing — we have sinned, and have committed iniquity — for himself and everyone before him, and asking one thing: that God look on the desolate city, for thy great mercies. He asks for no timetable.

The answer arrives before he finishes. Gabriel, sent at the beginning of thy supplications, calls Daniel greatly beloved and hands him a span running centuries past his own death. A man on his knees in a foreign capital asks for mercy on a ruin and is given seventy sevens. That mismatch is the strangest thing in the chapter, and the first thing a diagram removes.

The restraint about timing in those verses is not peculiar to Daniel, and when will Jesus return takes it up from the New Testament side. So when you go back to the chapter this week, give the first nineteen verses their share — the man who received the most calculated passage in the Old Testament was not calculating when it came.