It is hypothesized that procedural justice influences citizens' satisfaction with the police. An alternative argument holds that police performance measures, such as perceptions of crime and safety, are more salient. This study empirically investigates the predictive validity of both theoretical arguments. Using mail survey data from 563 adult residents from Monroe County, Michigan, a series of linear regression equations were estimated. The results suggest that procedural justice is a robust predictor of satisfaction with police. In contrast, several police performance measures failed to predict satisfaction with police. Overall, these findings support Tyler and Huo's (2002) contention that judgments regarding whether police exercise their authority in a procedurally-just fashion influence citizens' satisfaction with police more than fear of crime, perceptions of disorder, and the like.