This chapter examines the systematic factors behind local electoral results. Looking at data from over 7,000 different municipalities over a twenty-year time period, it appears that local elections are a curious mixture of the predictable and the idiosyncratic. They are predictable in that the majority of incumbents for local office either run unopposed or win reelection if they face challengers. This is consistent with the idea of managerial democracy: elections for local office should hinge on issues of custodial performance, and because incumbents get reelected at high rates, most are probably doing their jobs well enough to satisfy enough constituents or to dissuade any opponents. Identifying those instances when incumbents are likely to lose, however, turns out to be a very difficult task. Of the few identifiable trends, it appears that incumbent city council members are more likely to lose in places that are larger in size, greater in scope, and higher in bias. But the ability to predict the likelihood that any given incumbent is likely to lose, even when we know most political and social characteristics of a place, remains small.