AD 655

Muslim conquest of Khorasan

Arab retreat and continued raiding on the Khurasan–Tokharistan frontier

The date filed as a conquest may mark a retreat: Chinese records read by H. A. R. Gibb say Tokharistan's forces restored a Sasanian prince once the Arabs pulled back.

What 655 does not mark

No reliable scholarly chronology identifies 655 as the year Arab armies conquered both Khurasan and Transoxiana. Encyclopaedia Iranica dates the principal seizure of Nishapur and other Khurasani cities to 651–653, and stresses that the chronology varies.

H. A. R. Gibb's reconstruction from Chinese records instead places an Arab retreat in 655, after which forces from Tokharistan restored Yazdegerd III's son Peroz as a titular Persian king.

That specific date rests on Gibb's interpretation of those records; other accounts distribute the retreat, the revolts and the Arab counterattacks across 653–658.

Thin garrisons, repeated revolt

Abd Allah ibn Amir had withdrawn most of his army after imposing tribute. A reported 4,000 men were left to hold Merv.

Local forces then withheld tribute, expelled Muslims from Nishapur and recovered Balkh. Abd Allah ibn Khazim suppressed Qarin's revolt in 653–654, but Muslim control remained thin.

One modern scholarly summary describes the pattern as capture, rebellion and recapture.

Transoxiana, decades later

Reaching the Oxus or raiding Tokharistan was not the same as conquering Transoxiana, and that conquest was not completed in 655. The first traditionally dated Arab appearance at Bukhara was in 673–674.

Permanent Arab control there was achieved only during Qutayba ibn Muslim's campaigns of 706–709.

In the nearer term Arab authority collapsed across much of Khurasan during the First Fitna of 656–661.

Sources

Researched 24 Aug 2026 6 sources not yet audited Date disputed

How this was checked

Researched from the web into a fact sheet and rewritten from that sheet. This article has not yet had an independent model audit. How the pipeline works →

Date disputed. Sources disagree about when this happened.

What the sources leave uncertain

The eastern-conquest narratives contain duplicate campaigns, divergent Islamic-era dates and later tribal traditions; Encyclopaedia Iranica explicitly says the chronology varies.
Gibb's specific 655 retreat relies on interpretation of Chinese records. Other accounts distribute the retreat, revolts and Arab counterattacks across 653–658.
Because no discrete conquest of Khurasan and the Transoxianan Silk Road can be verified in 655, the listed event should not retain a solid-confidence date.