Forms of clitic doubling are attested in a significant number of Arabic varieties, including the Levant and northern Iraq, parts of Algeria and Morocco, Malta, Central Asia, and even, doubtfully, Dhofar. Language contact is widely accepted as an explanation for its presence in the Levant, and has been advanced as an explanation for its occurrence in North Africa, Malta, and Central Asia. However, none of the contact explanations proposed have yet addressed this phenomenon’s overall distribution across all of Arabic, usually limiting themselves to one or two regions at a time, and few have examined the parameters along which the relevant constructions vary. Without such an overview, it is not possible to exclude the hypothesis that clitic doubling simply reflects Arabic-internal trends, nor to determine whether its distribution reflects a single innovation or multiple independent ones. Updating the pioneering work of A. FISCHER (1907; 1909), this article demonstrates that clitic doubling has arisen independently within Arabic at least four times under the influence of different substrata/adstrata, and suggests areas in which more data on this construction would be especially useful.