Daniel 9:24–27 · Featured Study

The Seventy Weeks

Daniel is reading Jeremiah. He counts seventy years of exile, and prays. Gabriel interrupts the prayer with a larger number: seventy sevens are decreed, not for the ruins of a city but for the end of sin and the bringing in of everlasting righteousness. Then the angel breaks the number into segments and starts a clock. Every argument about this passage turns on two questions: when does the clock start, and does it ever stop.

Edward Poynter, Daniel's Prayer (1865) — Public domain, via Wikimedia Commons

Summary

Gabriel answers Daniel's prayer with a measured count: seventy sevens, decreed for the end of sin and the bringing in of everlasting righteousness. Everything argued about this passage turns on when the clock starts and whether it ever stops.

The six goals of verse 24 are about sin and righteousness, not a calendar.
Nearly every reader takes a 'seven' as seven years. The page shows the four reasons why.
Four royal decrees, almost a century apart, compete to start the clock.
The gap question: does the seventieth seven follow at once, or wait? Both sides are given.
Three computations land the count in the first century, and astronomy tests all three.

The Occasion

A prophet reading Scripture, and an answer larger than the prayer.

In the first year of Darius, Daniel says, he “understood by the books” the number of years Jeremiah had named for the desolations of Jerusalem: seventy. So he sets his face to seek God with fasting and confession. He is not asking for a prophecy. He is asking for his people. The answer multiplies his seventy by seven and hands him a calendar that runs past the rebuilt city altogether, to the Anointed One.

This is the greatest messianic disclosure in the book, and it comes to a man on his knees in the Scriptures.

The seventy that answers a seventy

Jeremiah’s seventy years of exile (Jer 25:11–12; 29:10) are answered by seventy sevens decreed on the same people and the same city. The echo is deliberate: the exile’s number becomes the measure of a far longer redemption.

Jeremiah 25 →

The Passage

Four verses. Read them before anybody counts anything.

Daniel 9:24 →
Seventy weeks are decreed upon thy people and upon thy holy city, to finish transgression, and to make an end of sins, and to make reconciliation for iniquity, and to bring in everlasting righteousness, and to seal up vision and prophecy, and to anoint the most holy.
The whole span and what it is for: six goals, all about sin and righteousness, none about a calendar.
Daniel 9:25 →
Know therefore and discern, that from the going forth of the commandment to restore and to build Jerusalem unto the anointed one, the prince, shall be seven weeks, and threescore and two weeks: it shall be built again, with street and moat, even in troublous times.
The start line and the first two segments. Hidden inside the punctuation is the quietest disputed decision in the passage.
Daniel 9:26 →
And after the threescore and two weeks shall the anointed one be cut off, and shall have nothing: and the people of the prince that shall come shall destroy the city and the sanctuary; and the end thereof shall be with a flood, and even unto the end shall be war; desolations are determined.
Two events placed after the sixty-two: an Anointed One cut off, and a city destroyed. The word after is why anyone argues for a gap.
Daniel 9:27 →
And he shall make a firm covenant with many for one week: and in the midst of the week he shall cause the sacrifice and the oblation to cease; and upon the wing of abominations shall come one that maketh desolate; and even unto the full end, and that determined, shall wrath be poured out upon the desolate.
The seventieth week alone, with an unnamed he at the front of it. Everything turns on who that pronoun is.

The Six Decreed Goals

Verse 24: what the whole count is for.

sin dealt with

1
To finish transgression
The rebellion is brought to its end. Not managed; finished.
2
To make an end of sins
The sin account is sealed up and closed out.
3
To make reconciliation for iniquity
Atonement: the covering that answers guilt. The verb is the language of the Day of Atonement.

righteousness brought in

4
To bring in everlasting righteousness
Not a reform that decays: a righteousness of the ages, brought in from outside.
5
To seal up vision and prophecy
The prophetic word reaches its term, sealed because fulfilled.
6
To anoint the most holy
A most-holy thing, or place, or one, consecrated at the close of the count.

Three goals close an account and three open one. Whatever else the count measures, it is measuring the distance to an atonement. That is why most Christian readings of the passage, Calvin’s and Anderson’s included, land on the cross somewhere inside it.

Why a Seven Is Seven Years

The step every computation takes in its first move.

The passage never says “years.” It says seventy sevens. Every computation on this page, Calvin’s and Anderson’s alike, turns that into 490 years in the first move. Here are the four reasons.

Daniel was already counting years

The whole scene starts because he read a number of years in Jeremiah and did the arithmetic. Gabriel answers in the unit Daniel brought.

Daniel 9:2 — “I, Daniel, understood by the books the number of the years … even seventy years.”

The Hebrew word is “a seven,” not “a week”

Shabua names a group of seven of anything. When Daniel means seven days he says so: in 10:2 the Hebrew reads literally three sevens of days, and the ASV prints “three whole weeks.” That extra word is doing real work, and 9:24–27 has no such word attached.

Daniel 10:2 — “In those days I, Daniel, was mourning three whole weeks.”

Israel already counted in sevens of years

Every seventh year the land itself kept a sabbath. The exile is then explained as the land collecting the sabbaths it was owed. That turns seventy years of exile into a ledger of seventy missed sabbath years, that is, 490 years of covenant history.

2 Chronicles 36:21 — “until the land had enjoyed its sabbaths … to fulfil threescore and ten years.” (Leviticus 25:4)

490 days will not hold the content

Sixteen months cannot contain a ruined city rebuilt “with street and moat, even in troublous times,” an Anointed One cut off, and a second destruction of city and sanctuary. The events demand a span of years.

Daniel 9:25–26

That much is nearly common ground; the critical reading counts years too. The argument starts one step later, at a question the passage also declines to answer: how long is a year?

The case for a 360-day prophetic yearGenesis 7–8 · Daniel 12 · Revelation 11–13

Genesis dates the flood to the day. The waters prevail 150 days (7:24) across an interval the narrative brackets from the seventeenth day of the second month (7:11) to the seventeenth day of the seventh month (8:4): five months of exactly thirty days.

Revelation supplies the day-count. It measures one span three ways: 1,260 days (11:3; 12:6), forty-two months (11:2; 13:5), and “a time, and times, and half a time” (12:14). Those are the same length only if a month is thirty days and a year is 360. Daniel uses the same “time, times, and a half” phrase (7:25; 12:7) but never attaches a day-count to it, so the 360-day year is read back into Daniel from Revelation rather than stated by both.

That is the arithmetic underneath Anderson’s 173,880 days.

The case against leaning your date on itthe counterweight

Israel did not live on a 360-day year. The kept calendar was lunar and solar together: twelve months of roughly 354 days, with a thirteenth month inserted often enough to keep Passover in the spring. A 360-day year is a prophetic convention inferred from other books; Daniel 9 never states a year length at all.

Daniel’s own day-figures do not come out to 360-day years either. The numbers he actually gives are 1,290 days (12:11) and 1,335 days (12:12). Neither divides into the 1,260 that the convention rests on.

The adjustment is also not one method but a family of them. Julius Africanus, running this same passage in the third century, reached a first-century landing using 354-day lunar years. Africanus and Anderson cannot both have the intended unit, so any day-exact result should be held loosely.

Variable One: Which Decree Starts the Clock

9:25 dates the count from the going forth of the commandment.

Decree	Date	What it authorized	Who starts here
CyrusEzra 1:1–4 →	538 BC	The return from exile and the rebuilding of the house of God at Jerusalem.	Used by readers who take the count as schematic rather than exact; the classic objection is that it commissions a temple, not a city.
Darius IEzra 6:1–12 →	c. 519 BC	A confirmation and funding of the Cyrus decree after the work had stalled.	Rarely made the starting gun; it renews an existing authorization rather than issuing a new one.
Artaxerxes, 7th yearEzra 7:11–26 →	458 / 457 BC	Ezra sent with authority over temple, treasury, magistrates, and the law of his God.	The anchor of the computations that land the 69 sevens near the beginning of Christ's public ministry. Two dates because two calendars: counted spring to spring, as the Persian court did, the seventh year is 458 BC; counted autumn to autumn, as Jewish civil reckoning did, it is 457 BC. The 457 track below needs the autumn figure.
Artaxerxes, 20th yearNehemiah 2:1–8 →	445/444 BC	Nehemiah sent to rebuild the city itself: the wall and the gates.	The anchor of the dispensational computation, on the ground that 9:25 speaks of a command to restore and build Jerusalem, a city, not a temple.

Why it matters: the four candidates sit almost a century apart. Move the starting gun from 538 to 445 and the end of the sixty-ninth seven moves with it by ninety-odd years. That is why two careful readers, using the same arithmetic, can land on two different decades of the first century.

Where the Sentence Breaks in 9:25

Your English Bible has already made a decision here.

Verse 25 names two numbers, seven and sixty-two. Whether they are one span running to one Anointed One, or two spans running to two different anointed figures, is settled by a pause. Translations do not all put the pause in the same place.

One span: sixty-nine to Messiah

“…from the going forth of the commandment to restore and to build Jerusalem unto the anointed one, the prince, shall be seven weeks, and threescore and two weeks…”

ASV, and with it the KJV, ESV, NASB and NIV

The two numbers are read together: 49 + 434 = 483 years to one Anointed One, with the seven sevens marking the rebuilding inside that longer stretch. This is the reading the diagram below draws.

Two spans: two anointed ones

The RSV and NRSV close the sentence after the first number: to the coming of an anointed one, a prince, there shall be seven weeks. Then for sixty-two weeks it shall be built again.

Now there are two figures: someone anointed 49 years after the decree (Cyrus, whom Isaiah 45:1 calls God’s anointed, or Joshua the high priest), and a second figure cut off 434 years later. That second figure is where the Maccabean reading puts the murdered high priest Onias III, in 171 BC. The critical case for the split is made at length by John Collins in his Daniel commentary (Hermeneia, 1993).

What actually decides it: a mark under the Hebrewthe athnach at 9:25

The consonants of the Hebrew carry no punctuation. The pauses were added centuries later by the Masoretes, the Jewish scribes who fixed the vowels and the chanting accents between roughly AD 600 and 1000. Their strongest mid-verse pause is called the athnach, and in Daniel 9:25 it stands under “seven weeks.”

Follow that accent and the sentence divides after the first number, which is the RSV and NRSV reading. Read the numbers as one series and you get the continuous sixty-nine. Both are defensible from the consonants alone; neither is invented.

The accents are an old and valuable reading tradition, but they are a tradition of reading, not the text itself. The ancient Greek versions of Daniel, centuries older than the accents, run the two numbers together toward one anointed prince. Some translations record the decision in a footnote at this verse; most say nothing.

The Count

Seven, then sixty-two, then one, and a question in between.

7 sevens

49 years

The street and the moat built again, in troublous times (9:25).

62 sevens

434 years

The long stretch running to the anointed one, the prince (9:25).

the gap question

Between the 69th and the 70th: nothing, or an age?

the 70th seven

7 years

A covenant confirmed with many, broken at the midpoint (9:27).

The final seven is drawn wider than its share so it can be read; the sixty-two sevens carry the length. The dashed interval is drawn because readers disagree about whether it exists, not because the text names it.

Variable Two: Does the Clock Stop

The gap question, both sides.

Yes: the seventieth seven waits

Verse 26 places the cutting off of the Anointed One and the destruction of city and sanctuary afterthe sixty-two sevens, and only then does verse 27 speak of the covenant of one seven. On this reading the text itself steps outside the count between the sixty-ninth and the seventieth. Jesus, generations after AD 70 would have closed the matter, still speaks of Daniel’s abomination as future (Matt 24:15), which requires something in Daniel 9 still outstanding.

No: the count runs straight through

If the prophecy says seventy sevens, the plain reading is seventy sevens straight through: not sixty-nine, then an unstated pause, then one more. The natural subject of “he shall make a firm covenant” in verse 27 is the Anointed One of verse 26, and His death is exactly what causes sacrifice and oblation to cease. On this reading the seventieth seven is the generation of Christ and the apostles, and the destroying prince is Rome under Titus.

Older Than the Argument

Two third-century readers of this passage.

Two charges get made about this passage: that counting the sevens to a dated Messiah is a Victorian parlour trick, and that the waiting seventieth week was invented in the nineteenth century. Both are answerable from the third century.

c. AD 221 · the count

Julius Africanus

The earliest surviving Christian computation of the seventy sevens is not Anderson’s. Africanus, assembling his world chronicle around AD 221, ran the count from Artaxerxes’ twentieth year to Christ. Finding that ordinary years overshot, he treated the 490 as Hebrew lunar years of 354 days, which he equated with 475 Roman years.

The move Anderson is famous for, counting the sevens as years and then adjusting the length of the year, is his by seventeen centuries. Only fragments survive, preserved by Eusebius.

What it settles: computing the sevens to Messiah is patristic practice, not a modern novelty. What it does not: which year-length is right. Africanus used 354 days and Anderson 360; they cannot both be the intended unit.

Julius Africanus, Chronographiae (c. AD 221), surviving in Eusebius, Demonstratio Evangelica 8

c. AD 204 · the gap

Hippolytus of Rome

The oldest surviving Christian commentary on any book of the Bible is Hippolytus on Daniel, written about AD 204. It already detaches the seventieth week. He reads the sixty-nine as running to Christ and the last seven as still outstanding at the end of the age, split at its midpoint: the two witnesses of Revelation 11 prophesying 1,260 days in the first half, the persecutor in the second.

That is the dashed interval in the diagram above, drawn more than 1,600 years before John Nelson Darby was born.

What it settles: the detached seventieth week is patristic. What it does not: Hippolytus is no dispensationalist. He has no church-age parenthesis, no distinct programme for Israel and the church, no pretribulation rapture. The system is modern; this one reading inside it is not.

Hippolytus, Commentary on Daniel 4.35–51 (c. AD 204)

The two systems, side by side →

The Six Readings

Each reading, with its strength and its weak point.

Dispensational-futurist

This site's tradition

Sir Robert Anderson, The Coming Prince (1894); Chuck Missler, The Book of Daniel

The 69 sevens run from Artaxerxes' 445 BC decree to Messiah's presentation at Jerusalem. The Anointed One is then cut off after the 69th: the cross falls between the sevens, not inside the 70th. A prophetic parenthesis, the church age, separates the 69th from the 70th; the 70th seven is a still-future seven years, and the one who confirms and then breaks the covenant in 9:27 is the coming ruler, not the Messiah.

Its strength: Takes 9:26's sequence at face value: the Anointed One is cut off after the sixty-two sevens, and the destruction of city and sanctuary follows. It also gives a natural referent to the covenant-breaker who sets up the abomination Jesus still treats as future (Matt 24:15).

Where it strains: It must locate a gap the text does not name, and the gap has now run far longer than the whole count it interrupts.

Continuous-messianic

John Calvin, Commentaries on Daniel; much of the Reformed tradition after him

No gap. The seventy sevens run unbroken to Christ. The Anointed One is cut off in the middle of the seventieth seven, and it is Messiah, not an antichrist, who confirms the covenant with many; His death causes sacrifice and oblation to cease by fulfilling them. The prince who is to come, whose people destroy the city and the sanctuary, is Rome under Titus in AD 70.

Its strength: Reads the covenant-confirmer of 9:27 Christologically, which is the most natural antecedent in a passage about the Anointed One, and needs no two-thousand-year parenthesis.

Where it strains: The arithmetic from any decree to the crucifixion strains to land the cross at the exact midpoint of a final seven.

Historicist (year-day)

Sir Isaac Newton, Observations upon the Prophecies of Daniel (1733); the Adventist year-day tradition after him

One prophetic day stands for one year, and the seventy sevens are a continuous 490-year block, read as the opening portion of the 2,300 days of Daniel 8:14, running from Artaxerxes' seventh-year decree (Ezra 7, 457 BC) to a first-century terminus at the Messiah's baptism and death. Nothing inside the seventy weeks is deferred. It is Daniel 7 and Revelation, not this passage, that historicists unroll across the long course of the church age.

Its strength: Preserves an unbroken chronology, ties Daniel 9 back to the vision of Daniel 8 as its own explanation, and applies one year-day rule consistently across both books.

Where it strains: Both the 457 BC starting point and the year-day rule are brought to the passage rather than read out of it, and the wider historicist scheme's later identifications have repeatedly needed revision as the centuries it assigned came and went.

Maccabean (the critical reading)

The standard reading in critical scholarship — e.g. John J. Collins, Daniel (Hermeneia, 1993)

The clock terminates at Antiochus IV. The anointed one cut off in 9:26 is the murdered high priest Onias III (171 BC); the covenant-breaker is Antiochus; the 490 is schematic rather than chronometric.

Its strength: Fits the acknowledged Antiochus focus of chapters 8 and 11 and needs no figure outside the book's own historical horizon.

Where it strains: The arithmetic from any of the candidate decrees to 164 BC does not cleanly yield 490 years, and the reading depends on the second-century dating that is itself contested.

Reformed partial-preterist (AD 70)

R. C. Sproul, The Last Days According to Jesus (1998)

The last-days prophecies were fulfilled in the first century, culminating in the destruction of Jerusalem in AD 70: a datable, historical landing. On this reading the anointed one cut off is Christ, and the prince whose people destroy city and sanctuary is Rome; the judgment Jesus announces in Matthew 24 falls within the generation He addressed. Sproul holds the return of Christ itself still future.

Its strength: Takes seriously that Jesus tied a great deal to 'this generation' (Matt 24:34), and gives 9:26's destruction of the city a fulfillment inside recorded history.

Where it strains: It has to explain what remains for the seventieth seven once AD 70 has absorbed the weight of the prophecy, and how far 'fulfilled' can be pressed without emptying the final consummation.

Symbolic

Keil and Delitzsch, Commentary on Daniel; the broader Reformed amillennial stream

The numbers are theological rather than chronometric: seventy sevens as consummate completeness, in the same register as the seventy times seven of Matthew 18:22. Keil declines any arithmetical computation of the sevens and places the seventieth week not in the first century but at the close of the present course of the world, with the Antichrist destroyed inside it.

Its strength: Honors the caution against turning prophecy into arithmetic, and refuses to make the passage's authority depend on a calculation.

Where it strains: The passage plainly counts, in segments of seven and sixty-two, and a purely symbolic reading has to account for why it bothers.

Three Computations, Run Side by Side

Each with its assumptions and its costs shown.

People usually meet one of these and think it is the computation. Three are in common use. They disagree about the decree, the length of a year, and the landing.

Anderson

1894This site's tradition

Sir Robert Anderson, The Coming Prince (1894), public domain

Starts at: Artaxerxes' 20th year: Nehemiah 2, 445 BC
Year length: 360 days, the prophetic year
Span: 69 × 7 = 483 years = 173,880 days
Runs: 14 March 445 BC → 6 April AD 32
Lands on: the entry into Jerusalem, days before the cross

What it costs: AD 32 cannot carry a Friday crucifixion at all (see the astronomy below), and Anderson's Julian date for the Persian new year of 445 BC does not match the reconstructed calendar. His arithmetic is checkable; his two anchors are the part that failed.

Hoehner

1977

Harold Hoehner, Chronological Aspects of the Life of Christ (1977), ch. 4, cited, not quoted

Starts at: the same decree, dated a year later: 444 BC
Year length: 360 days, unchanged
Span: the same 173,880 days
Runs: 5 March 444 BC → 30 March AD 33
Lands on: the entry into Jerusalem, with the cross that Friday, 3 April AD 33

What it costs: It still needs the 360-day year, which Daniel 9 never states, and it still rests on a decree date reconstructed from Persian regnal reckoning rather than read off the page.

The 457 track

the older continuous reading

the continuous-messianic and historicist stream, anchored in Ezra 7 rather than Nehemiah 2

Starts at: Artaxerxes' 7th year: Ezra 7, autumn 457 BC
Year length: ordinary years, no conversion at all
Span: 483 years, counted straight
Runs: autumn 457 BC → autumn AD 27
Lands on: the anointing at the Jordan (Luke 3:21–22); then “the midst of the week,” the cross; the 490 closing about AD 34

What it costs: It needs the crucifixion about AD 31, and AD 31 is not one of the years in which Passover eve fell on a Friday. It also has to hear “the commandment to restore and to build Jerusalem” in a decree that commissions a temple, a treasury, and a magistracy.

Anderson published his working, which is why it can be checked at all. Checking it is what produced Hoehner’s revision. Both are computations, not proofs: the dates at both ends are reconstructions, and the exercise presupposes the sixth-century dating of Daniel.

The One Test From Outside the Argument

Which years could carry a Friday crucifixion.

All three tracks land at a Passover. That makes the crucifixion year testable without reference to Daniel at all: rebuild the first-century Jewish calendar from the moon and ask which years put the fourteenth of Nisan (Passover eve, the day the Gospels place the cross) on a Friday.

Colin Humphreys and Graeme Waddington did that calculation and published it in Nature in 1983, working from lunar visibility rather than from any prophecy. Within the years Pilate held office, only two qualify: Friday 7 April AD 30 and Friday 3 April AD 33. They argued for AD 33.

Harold Hoehner had reached the same year six years earlier by a different road, and it is the reason his revision exists: AD 32, Anderson’s landing year, has no Friday Passover eve to offer, so Anderson’s dates cannot stand as he published them. The same test removes about AD 31, which the 457 track needs.

What survives the test

AD 33:Hoehner’s landing, and Humphreys and Waddington’s preferred year.
AD 30: still fully live, and held by many careful chronologists.
AD 31, AD 32: ruled out by the calendar, which costs Anderson his landing and the 457 track its usual one.

The remainder

If the crucifixion was AD 30, none of the three landings above is right. Astronomy narrows the field to two years; it does not hand anyone the passage.

Harold Hoehner, Chronological Aspects of the Life of Christ (1977), ch. 4 · Colin Humphreys and Graeme Waddington, “Dating the Crucifixion,” Nature 306 (1983): 743–746.

The Keynote Words

The vocabulary of the count.

שָׁבוּעַ

shabua

a seven; seventy of them, read by nearly all as sevens of years

מָשִׁיחַ

mashiach

Anointed One; 9:25 and 9:26, the only two occurrences in a datable countdown

Word studies

The Hebrew and Greek under this study, each word in its own room.

Word roomshabua, a seven→Word roommashiach, Anointed→

Keep studying

Read it5

By subject

Typology27 pages

All 27 pages on Typology

Jesus85 pages

All 85 pages on Jesus

The whole shelf: Christ Hidden in the Old Testament →