This article provides an empirical test of the common assumption that public support for vigilantism is affected by confidence in police. Aside from assessing the role of diffuse (general) confidence in police, we also tested whether police response on a situational level affects how the public views an act of vigilantism. Respondents (N=385) were presented with a vignette about vigilantism. Using an experimental between-subjects design, we varied police responsiveness (high/low) to precipitating crime as well as vigilante violence (high/low). Diffused confidence in police was a significant predictor of support for vigilantism. Additionally, both experimental factors played an important role: low police responsiveness and low vigilante violence led to more support for vigilantism. Citizens are thus sensitive to situational variation when judging a crime. Our findings also emphasise the importance of police action on a local level for the formation of public opinion. © 2013 © 2013 Taylor & Francis.