Roughly twenty authors describe earthquakes that caused damage across a vast region—the Hejaz, Ayla, Ramla, Jerusalem, Mesopotamia, parts of Egypt, and Banias. If all these reports are accurate, more than one earthquake must be invoked to explain such wide-ranging effects. The earliest testimony comes from the personal diary of Abu Ali ibn al-Banna of Baghdad. In August 1068 CE he recorded reading of two distinct shocks. The first occurred on Tuesday 18 March 1068 CE, destroying towns in the Hejaz, Ayla, Ramla, and Jerusalem; an unlocated tsunami was also mentioned. On that same day Baghdad felt “a slight earthquake.” The second, on Thursday 29 May 1068 CE, was reported by caravan merchants who claimed that Ramla was devastated—“all its dwellings except two” destroyed, 15 000 killed—with similar damage in Jerusalem and another unplaced tsunami. The destruction repeated that of the March account, raising doubts about duplication.

About a century later, the Baghdadi historian Ibn al‑Jawzi described an earthquake between 8 March and 6 April 1068 CE, naming many of the same localities as al-Banna and adding Ruhba and Kufa. Although he likely drew from al-Banna’s diary, he claimed to quote a letter written by merchants. Makdisi (1956a) argued that Ibn al-Jawzi may have possessed parts of al-Banna’s now mostly lost diary. About 180 years after the event, Sibt ibn al‑Jawzi — Ibn al-Jawzi's grandson — produced an expanded version of the 18 March earthquake, repeating many details and adding the overflowing of the Euphrates and destruction at Banias. Raised in his grandfather’s Baghdad household, he may have inherited his extensive library.

Another contemporary witness, Mawhub ibn Mansur ibn Mufarrij, writing in Egypt between 1088 and 1094 CE, based his account on a monk in the Nile Delta. He described a mid-morning shock on 18 March 1068 CE that ruined Ramla and obliterated Ayla. Jerusalem and Tinnis in the eastern Delta also suffered, along with an unlocated tsunami, damage to Cairo’s congregational mosque, and two aftershocks.

The 18 March date recurs in most sources, many naming Ramla as a chief casualty. Only al-Banna added 29 May, repeating the same effects. Whether that represents a second quake or duplication remains uncertain. Ramla’s losses may have been overstated: the town had been shattered in 1033 CE by the 11th-century Palestine Quakes, and al-Maqrîzî records a Turkish attack in Muharram A.H. 460 (11 Nov–10 Dec 1067 CE). Merchants reaching Baghdad in 1068 CE may have conflated earthquake and war damage. Built on loose sands with a shallow water table, Ramla is also prone to liquefaction. Reports of wells overflowing by multiple authors strongly indicate such effects, explaining the disproportionate destruction.

Some modern works place a tsunami in the Mediterranean, but no medieval author specified its location. It may instead have struck Ayla, where several writers say that only twelve fishermen survived. If true, their absence at sea suggests a daytime quake followed by a tsunami that engulfed the coast. Other accounts mention fissures and new springs in the Hejaz. The strongest paleoseismic and archaeoseismic evidence today comes from the southern Araba, indicating that faults broke there. Paleolandslide evidence also exists in the Gulf of Aqaba, explaining the damage reported in the Hejaz and probable tsunami in Ayla.