Alongside negative feedback loops between violence and development, several studies find a positive and significant effect of exposure to warfare onto prosocial behavior. This increase in prosociality tends to favor the group the individual identifies with relative to the out-group. The current study assesses whether these findings extend to victims of crime in urban areas. We conduct two artefactual field experiments in Bogotá (Colombia) to address these issues. Our methodological strategy is to experimentally manipulate the recall of violence, either through a direct question or through a monetary loss in participants’ experimental endowment. We interact these treatments with the degree of exposure to violence. We find that victims recalling experiences of urban violence act more prosocially in terms of trust, trustworthiness, and cooperation. The increase in prosociality favors residents in the same city district as the participant (ingroup bias). However, the ingroup bias holds in trust decisions but not in public goods games decisions. We fail to find statistically significant evidence supporting any mediation mechanisms we analyze.