The Book of Genesis reached its final form almost certainly between the sixth and fifth centuries BCE (roughly 540–400 BCE), but it was assembled from sources written centuries earlier, the oldest of which may date to the tenth or ninth century BCE. No single date is correct, because Genesis is not a single document. It is a layered compilation, and its timeline stretches across roughly 600 years.

What “written” actually means for Genesis

Genesis is anonymous in its own text. It never names an author, and the tradition attributing it to Moses emerges much later. That matters for dating, because the question “when was it written” splits into three separate questions: when were its underlying sources composed, when were those sources combined, and when did the text stabilize into the form we have now? Each has a different answer and a different set of evidence.

The text itself gives clues to its own formation. Genesis 12:6 notes that “the Canaanites were then in the land,” language that reads like a later editor explaining a vanished situation. Genesis 36:31 mentions kings of Israel “before any king reigned over the Israelites,” a remark that presupposes the monarchy already existed. Doublets, such as two creation accounts and two flood narratives with different details, signal combined traditions rather than one author writing from scratch. These internal seams are the raw material of every dating argument.

Linguistic analysis completes the picture. Genesis contains classical Biblical Hebrew mixed with features that resemble later, post-exilic Hebrew, including Persian loanwords elsewhere in the Hebrew Bible and Aramaic influence that increases over time. Statistical studies of Hebrew syntax place Genesis in a middle-to-late stratum, not the earliest. The language alone cannot fix a date, but it rules out any single-author model.

Plain version: the book looks like a library, not a letter.

The documentary hypothesis and its four proposed sources

The dominant framework for two centuries has been the documentary hypothesis, most associated with Julius Wellhausen in the 1870s. It proposes that Genesis combines four strands, conventionally labeled J (Yahwist), E (Elohist), P (Priestly), and D (Deuteronomic, mostly outside Genesis). The letters reflect characteristic names for God and distinctive vocabulary and theology. Later scholarship has softened the model considerably. Many European scholars now speak of a “Priestly” layer and older narrative traditions without insisting on four clean documents.

The proposed source dates follow a familiar sequence. J is often placed in the tenth or ninth century BCE, in the southern kingdom of Judah. E is placed in the ninth or eighth century, in the northern kingdom. P is placed in the sixth century, during or after the Babylonian exile. Final redaction, the combining of sources into one scroll, is placed in the Persian period, roughly 539–400 BCE. These are scholarly conventions, not measurements, and the ranges move when new arguments appear.

The Priestly material matters most for the final date. Its genealogies are structured with precise numbers and ages, its ritual interests align with later priestly concerns, and its covenant theology fits a community rebuilding after exile. That alignment does not prove the date, but it makes a sixth-century origin for the P strand the most economical explanation.

Quantifying the timeline: dates, ranges, and manuscript evidence

Dating claims in this field are ranges, and the honest way to present them is as ranges with reasoning attached. The table below summarizes the current state of play.

Element Conventional date range Main basis Confidence
Oldest poetic traditions (e.g., Genesis 49) 10th–9th c. BCE Archaic Hebrew features, parallel to Ugaritic poetry Moderate
J source (narrative strand) 10th–8th c. BCE Divine name usage, narrative style, Judahite setting Contested
E source 9th–8th c. BCE Elohim usage, northern settings, Joseph story Contested
P source (priestly layer) 6th–5th c. BCE Genealogical system, ritual concerns, exile setting Fairly strong
Final redaction 540–400 BCE Persian-period framing, combined seams Strong
Earliest complete manuscripts c. 1000–1100 CE (Masoretic) Cairo Codex, Aleppo Codex, Leningrad Codex Certain

The manuscript column needs emphasis. The oldest complete Hebrew manuscripts of Genesis date to about the tenth or eleventh century CE. The famous Dead Sea Scrolls push fragmentary evidence back to roughly the third century BCE, and those fragments show textual variation, not uniformity. So the gap between composition and our earliest physical copies is enormous, exceeding a thousand years. That gap is why internal evidence, language, and comparative ancient Near Eastern texts carry the dating burden.

External comparison sharpens the numbers. The Babylonian Epic of Gilgamesh includes a flood narrative with striking parallels to Genesis 6–9, and its standard version dates to roughly the seventh century BCE, though older versions go back further. Egyptian and Mesopotamian creation stories offer similar structural parallels. Genesis shares literary furniture with its neighbors, which supports a first-millennium setting for the final text regardless of how old its traditions are.

The professional caveat: why the four-source model is under pressure

The counterintuitive detail is this: the documentary hypothesis is now more contested than any popular summary admits, yet the late date of Genesis's final form is among the least contested conclusions in the field. Those two facts coexist because the critique targets the source model, not the timeline.

Scholars such as Rolf Rendtorff and later Thomas Römer argue that the sources were never independent, complete documents. Instead, they describe a long process of fragmentary traditions, small collections, and successive expansions, with the Priestly writers functioning as editors as much as authors. This model predicts more unevenness, more redundancy, and more editorial intrusion than the tidy four-document picture. It also weakens attempts to date J precisely, since J may be a retrojected label for material that accumulated for centuries.

A second pressure point is the Persian-period hypothesis. Some scholars, including for a time Peter Frei's work on imperial authorization, proposed that the Torah was assembled under Persian sponsorship around 450–400 BCE. That specific administrative theory has been criticized, but the broader Persian-period window for final redaction remains mainstream. The practical upshot for readers: treat “when was the book of Genesis written” as a question about a process, and treat any single year as a midpoint of an argument, not a fact.

Watch for these patterns when reading secondary sources on Genesis:

A single precise year presented without a range, such as “Genesis was written in 1445 BCE.” That date derives from internal biblical chronologies, not from archaeological or linguistic evidence. Claims that the Dead Sea Scrolls prove the text was unchanged before 200 BCE. The scrolls show textual plurality, and the standardization came later. Arguments that treat the documentary hypothesis as either proven fact or total fraud. Both overstate what the evidence can bear. Citations of “Hebrew word studies” as decisive when the argument rests on one word's distribution in a text of debated date. Dates for J that predate the monarchy, often defended on theological rather than linguistic grounds.

The safest red flag, in practice, is certainty itself. The field's own journals publish competing dates for the same stratum within the same decade.

Practical steps: how to evaluate a claim about Genesis's date

If you want to assess any specific dating claim, work through these steps in order.

Identify what is being dated. Is it the oldest tradition, a specific source, the final redaction, or the manuscript? Confusing these is the most common error. Check the evidence type. Linguistic features, ancient Near Eastern parallels, and internal cross-references carry different weights, and the claim should say which it uses. Ask for the range, not the point. A responsible claim gives a century-scale range and states its confidence level. Compare against the manuscript ceiling. Nothing about composition can be tested directly against a tenth-century CE manuscript, so the argument must rest on internal and comparative evidence. Distinguish evidence from inference. The P layer's exile setting is an inference from content; the sixth-century range is a conclusion built on that inference.

A concrete threshold helps: if a source gives your date to the year, downgrade its reliability. If it gives a range wider than a century and explains the reasoning, it is doing the work correctly.

What the current evidence cannot settle

Several genuine limits remain. First, archaeology rarely dates texts directly; it dates destruction layers and settlement patterns, which at best bracket cultural settings. Second, oral tradition leaves no physical trace, so the pre-history of any Genesis story is effectively unknowable. Third, the Hebrew Bible's own chronology and the ancient Near Eastern chronology each carry their own uncertainties, and reconciling them introduces error.

The strongest defensible summary is this: the traditions in Genesis include very old material, the Priestly layer and the final redaction belong to the sixth and fifth centuries BCE, and the book reached its stable form sometime before the third century BCE, since the Dead Sea Scrolls already show it as an authoritative text with variant readings. Beyond that, the dates are arguments. The honest answer to when the Book of Genesis was written is a span measured in centuries, anchored at its late end by the Persian period and at its early end by traditions older than the monarchy.