Absent from the literature on mayoral control of urban public schools are quantitative studies that create a coherent theory of why mayoral control occurs. This article attempts to fill this gap by analyzing 47 urban school districts from 1991 to 2002, using a Cox Proportional Hazards model to assess the effects of political, economic, and district demographic variables on the risk of adopting mayoral control. The author finds that districts in states with Republican legislatures and large impoverished populations are most likely to adopt mayoral control. Lastly, over time, large union membership actually increases the likelihood of adoption, contrary to the conventional wisdom that teachers' unions obstruct education reform. (Contains 5 tables, 2 figures, and 13 notes.)