This PhD thesis redefines the practice of siege warfare in Syria and the Jazīra, from 1097 to 1192, in two ways. First, it uses an innovative quantitative approach to historical narratives. The systematic exploitation of every siege event mentioned in 51 narratives provides the largest dataset of medieval sieges gathered to this day, with 730 sieges. Therefore, the norm has been studied rather than the exceptional, leading to four major benefits: a clear and detailed description of the model siege operation; illumination of hitherto overlooked aspects of poliorcetics such as logistics and sallies; the revaluation of key claims regarding siege warfare in the scholarship; the redefinition of key phases and strategies in the twelfth century. We can outline a model siege: it occurred in inland Syria between Muslim besiegers and either Frankish or Muslim besieged. The model siege was successful for the besiegers and ended after a maximum of forty-five days, by either an assault or a surrender. For the first time, besieged logistics are discussed, revealing the existence of a standard operating procedure in gathering supplies and a remarkable ability of medieval commanders to provide their settlements with adequate supplies. This thesis also explores the too often overlooked question of sallies, showing their benefits and the competent handling of their risks by the besieged. The thesis has also re-evaluated the existence and the chronology of both Muslim and Frankish strategies in the period. Second, it gives equal treatment to Latin and non-Latin sources, the latter evidence being too often rejected by favouring Latin evidence. The near omnipresence of Muslim belligerents provides historians with the clear message that to understand warfare in the twelfth century we need to adjust our approach by giving our primary focus to Muslim belligerents, rather than following the historiographical tradition of focusing on the Franks.