When Was the Dome of the Rock Completed — 691 or 692?

The Dome of the Rock carries its own date: an inscription in the inner octagonal arcade names the year AH 72, which runs from about June 691 to May 692 CE by standard tabular conversion. That single Hijri year straddles two Gregorian years, which is the whole reason reference works print "691," "692," or "691/692" for the same building — for the most part they are not disagreeing about the evidence, only about how to convert it. A separate and genuine question, taken up below, is whether AH 72 dates the completion at all. Britannica notes that the inscription dates the building's construction to AH 72 (691–692 CE), which is generally taken as the completion date, under the Umayyad caliph Abd al-Malik ibn Marwan (r. 685–705). What is genuinely debated is not the reading of the number but what AH 72 commemorates: the finishing of the building, or an earlier stage of the work.

Why one date produces two years

The Islamic calendar is lunar, and its years are about eleven days shorter than solar ones, so a Hijri year almost never lines up with a Gregorian year. AH 72 ran from June 691 CE to May 692 CE. Conversion tables place its start in the first days of June 691 and its end in the second half of May 692, differing among themselves by a day or two, because lunar-calendar conversion depends on which tabular scheme is used. An event dated only to "the year 72" therefore falls somewhere inside a window that spans the back half of 691 and the front half of 692.

This is why a machine comparing sources sees a contradiction where historians see none. A writer who converts AH 72 to its opening year prints 691; a writer who gives the full span prints 691/692; a writer following a source that placed the event in the later months prints 692. All three are converting the same inscription. If you need a single figure, "AH 72 (691–692 CE)" is the honest one, and it is what careful catalogue entries use.

What the inscription actually says — and the al-Ma'mun tell

Running around both faces of the inner octagonal arcade is a mosaic band of gold Kufic letters roughly 240 metres long. It is among the earliest securely dated monumental inscriptions containing Qur'anic material, and the earliest of any scale that survives intact, which makes it a first-rank document for the history of the Qur'anic text as well as for the building. Its content is pointedly theological: it addresses the People of the Book, insists that God has no son and no partner, and honours Jesus as a messenger and servant of God — language read by most scholars as a deliberate statement aimed at the Christian population of a recently conquered Jerusalem.

The inscription's most famous feature is a visible alteration. The name of the patron was later replaced with that of the Abbasid caliph al-Ma'mun (r. 813–833), whose reign began more than a century after the mosaic was laid — but the date was left untouched. The mismatch is self-refuting: al-Ma'mun's name sits beside a year he was not alive in. That clumsy edit is a large part of why the AH 72 date is trusted: it survives as the original craftsmen set it, and it independently corroborates the attribution to Abd al-Malik preserved in the later Arabic literary sources.

Abd al-Malik, Arabic, and the coinage reform

The Dome belongs to a programme, not a whim. Abd al-Malik emerged from the second civil war as the consolidator of Umayyad rule, and within a few years of the Dome he pushed through changes that made the caliphate's authority visible in everyday hands. From AH 77 (696–697 CE) his reformed gold dinars abandoned figural imagery altogether and carried purely Arabic religious inscriptions; reformed silver dirhams followed within a year or two. In roughly the same period the administrative registers of the empire began to be converted from Greek and Persian to Arabic.

Read together, the building and the coins do the same work in different media: monumental Arabic script asserting an Islamic confession in a landscape and an economy that had been Byzantine within living memory. The Dome is the earliest surviving monument of that programme, which is a large part of why its date matters so much.

The site before and after: 638 to the Crusades

Jerusalem passed to the Rashidun caliphate in 638 CE (some accounts place the surrender in 637), nearly six centuries after the Roman destruction of the Second Temple in 70 CE had left the Temple Mount — the Haram al-Sharif — a platform without a temple; its condition and use across the Roman and Byzantine centuries are debated, but no major structure stood on it when the Dome was begun. The Dome was raised on that platform some fifty years later, over the outcrop of bedrock that gives the building its name.

When the First Crusade took Jerusalem in 1099, the Dome was converted into a Christian church known as the Templum Domini, the Temple of the Lord, and a cross was set above it; the nearby al-Aqsa Mosque became a royal palace and later quarters associated with the Knights Templar, who took their name from the site. After Saladin retook the city in 1187, the building was returned to Muslim use. Its exterior has been reworked repeatedly since — most conspicuously the sixteenth-century Ottoman tilework under Suleiman the Magnificent, which replaced the original outer mosaics, and twentieth-century recladding of the dome itself. The interior arcade inscription, however, is original, which is why the building can still date itself.

What most scholars accept, and what stays contested

• Accepted: The inscription in the inner octagonal arcade reads the year 72 of the Hijra, equivalent to a year running from about June 691 to May 692 CE; the building was commissioned under the Umayyad caliph Abd al-Malik ibn Marwan; the patron's name in the inscription was later replaced by that of al-Ma'mun while the original date was left in place.
• Mainstream: Historians and reference works including Britannica treat AH 72 / 691–692 CE as the completion date of the Dome of the Rock, making it the oldest extant Islamic monument. Most of the apparent 691-versus-692 disagreement between sources is a calendar-conversion artefact rather than a dispute about the evidence; the separate question of whether AH 72 dates completion at all is a real one.
• Contested: Whether AH 72 marks completion or an earlier point in the project — the foundation, or the finishing of the structural shell before the mosaic decoration — has been argued, most prominently by Sheila Blair, who argued the date does not mark completion; on readings following her, the building would not have been finished until 75 AH or later, though no source gives a firm completion year. The building's purpose is also genuinely disputed: proposed readings include a victory monument over Byzantium, a theological statement addressed to Christians, and a commemoration of the sanctity of the Temple Mount; the once-popular claim that Abd al-Malik intended it as a substitute pilgrimage site rivalling Mecca is now widely rejected. The association with the Prophet's Night Journey is not stated in the original inscription and appears in the sources later.

Frequently asked questions

Explore related content on this site

• Construction of the Dome of the Rock
• The Early Muslim Conquests
• The Fall of the Second Temple
• The Islamic Golden Age

Sources and further reading

• Britannica's entry on the Dome of the Rock, dating the shrine to 691–692 CE under Abd al-Malik and describing it as the oldest extant Islamic monument.
• The foundational modern study of the building's inscriptions, iconography and likely purpose, which framed the scholarly debate that followed.
• The article arguing that the AH 72 date marks an earlier stage of construction rather than completion — the principal dissenting position on the date.
• The Metropolitan Museum of Art's overview of Umayyad art and architecture, covering the Dome of the Rock, Abd al-Malik's epigraphic coinage reform, and the turn to Arabic in administration.
• The standard epigraphic corpus recording and transcribing the Jerusalem inscriptions, including the arcade band and the substituted patron's name.