Much debate has taken place regarding the merits of aggressive policing strategies such as “stop, question, and frisk.” Labeling theory suggests that police contact may actually increase delinquency because youth who are stopped or arrested are excluded from conventional opportunities, adopt a deviant identity, and spend time with delinquent peers. But, few studies have examined the mechanisms through which police contact potentially enhances offending. The current study uses four waves of longitudinal data collected from middle‐school students (N = 2,127) in seven cities to examine the deviance amplification process. Outcomes are compared for youth with no police contact, those who were stopped by police, and those who were arrested. We use propensity score matching to control for preexisting differences among the three groups. Our findings indicate that compared with those with no contact, youth who are stopped or arrested report higher levels of future delinquency and that social bonds, deviant identity formation, and delinquent peers partially mediate the relationship between police contact and later offending. These findings suggest that programs targeted at reducing the negative consequences of police contact (i.e., poor academic achievement, deviant identity formation, and delinquent peer associations) might reduce the occurrence of secondary deviance.