How do citizens evaluate the performance of their mayors? Previous studies have examined mayoral performance either with cross-sectional surveys or by comparing pairs of consecutive elections. In this article, we use 150 surveys conducted in New York City between 1984 and 2009 to carry out the first time-series analysis of mayoral approval. We show that fluctuations in crime and the economy affect mayors’ ratings and that black and white citizens react similarly to changing local conditions (although their initial evaluations of mayors often diverge sharply). We also show that how New Yorkers rate mayors in the polls is closely related to how they vote for mayors at the polls.