R. Hoyland has characterised “the crux” of Patricia Crone's and Michael Cook's ‘alternative’ theory of the origins of Islam, which they call “Judaeo-Hagarism”, briefly and aptly as follows: “Arab-Jewish involvement in the conquest of their birthright of the Holy Land, united by common descent from Abraham and motivated by Judaic messianism”. For this theory, or rather hypothesis, the two authors rely mainly (but not exclusively; see below) on an account in a chronicle attributed to the Armenian bishop Sebeos (written around 655), which centres on the calling of an Ishmaelite (Arab) merchant and preacher, Mahmet (=Muḥammad), who was familiar with the Jewish religion, especially with the story of Moses. According to the proponents of the Hagarism hypothesis, Muḥammad called the united Ishmaelites and Israelites to conquer “their” land, Palestine, which was under Byzantine rule at the time. In the campaign against the Greeks (Byzantines) that soon followed, not only Ishmaelites but also 12000 Israelites are said to have taken part; the Ishmaelites are said to have distributed the Israelites among their tribes to use them as guides into the land of Israel. – Since this account contains the earliest information about the Islamic prophet and his career, proponents of the Hagarism hypothesis consider this “external evidence” to be the more likely alternative to the early Islamic reports of the origins of Islam; in fact, these reports, in the form we have today, were redacted much later. Crone and Cook in Hagarism argue as follows: The terms hiǧra and muhāǧirūn originally referred not to an alleged emigration of Muḥammad from Mekka to Medina, but to an exodus from Arabia to Palestine; it is only later, by a geographical emendation, that Palestine lost its centrality as the religious focus of Islam in favour of Mekka. The conquest of Palestine was led by Muḥammad himself in 634, thus two years after the generally accepted date of his death.

Before reviewing the most important texts on which Hagarism is based, it should be said in all fairness that Crone and Cook have gradually abandoned constituent components of the hypothesis underlying their book. Cook is now apparently critical of his and his co-author's methodology and working methods at that time. However, he does seem to continue to regard the texts to be discussed here as possible supports for his original hypothesis.