This study examines the effects of a 1-day pilot training program on ethnic/racial bias in police use of stop and search powers, using a randomized controlled trial in six diverse agencies in England. We theorized the training could reduce officer bias by improving their competence to apply legitimate criteria in search decisions, and/or by reducing their reliance on ethnic/racial stereotypes. Survey results showed the training improved officers' knowledge of stop and search regulations, made them more selective in declared search intentions in hypothetical scenarios, and reduced their support for ethnic/racial stereotyping in policing. While it showed no effects on the ethnic/racial patterns of search intentions in survey scenarios, there was no survey evidence of bias against black people in the scenarios, even in the absence of training. Police search records revealed no clear training effects on recorded street-level behaviors, whether in relation to the frequency of searches, the strength of grounds for suspicion, or their ethnic/racial patterning. Policy implications: A 1-day police training program to reduce ethnic/racial bias may change officers' knowledge and attitudes but, on its own, may not be sufficient to impact their street-level behaviors. Training is probably most effective as part of a package of reforms, and particularly when it: involves sufficient "dosage";