© 2015 American Psychological Association. Objective: We aim to use conditional or moderated mediation to simultaneously test how and for whom an injection risk intervention was efficacious at reducing receptive needle sharing among female sex workers who inject drugs (FSWs-IDUs) in Mexico. Methods: Secondary analysis of data from a randomized trial. A total of 300 FSW-IDUs participated in Mujer Mas Segura in Ciudad Juarez, Mexico, and were randomized to an interactive injection risk intervention or a didactic injection risk intervention. We measured safe injection self-efficacy as the hypothesized mediator and policing behaviors (being arrested and syringe confiscation) as hypothesized moderators. In total, 213 women provided complete data for the current analyses. Results: Conditional (moderated) mediation showed that the intervention affected receptive needle sharing through safe injection self-efficacy among women who experienced syringe confiscation. On average, police syringe confiscation was associated with lower safe injection self-efficacy (p =.04). Among those who experienced syringe confiscation, those who received the interactive (vs. didactic) intervention reported higher self-efficacy, which in turn predicted lower receptive needle sharing (p =.04). Conclusions: Whereas syringe confiscation by the police negatively affected safe injection self-efficacy and ultimately injection risk behavior, our interactive intervention helped to "buffer" this negative impact of police behavior on risky injection practices. The theory-based, active skills building elements included in the interactive condition, which were absent from the didactic condition, helped participants' self-efficacy for safer injection in the face of syringe confiscation.