BACKGROUND: Many individuals with obsessive–compulsive disorder (OCD) display behavioral avoidance related to their obsessional thoughts and compulsive behaviors. However, how these avoidance behaviors impact treatment outcomes with exposure and response prevention (EX/RP) remains unclear. We examined pretreatment avoidance behaviors as predictors of EX/RP outcomes. METHODS: Data came from a randomized controlled trial of augmentation strategies for inadequate response to serotonin reuptake inhibitors comparing EX/RP (N = 40), risperidone (N = 40), and placebo (N = 20). Baseline avoidance was rated with the avoidance item from the Yale–Brown Obsessive–Compulsive Scale (YBOCS). Primary analyses examined avoidance behaviors as predictors of EX/RP outcomes. To test specificity, we explored whether avoidance also related to outcomes among patients receiving risperidone and placebo. RESULTS: More than half (69%) of the full sample had moderate or severe avoidance behaviors at baseline. In EX/RP, controlling for baseline severity, pretreatment avoidance predicted posttreatment YBOCS symptoms (β = 0.45, P < .01). Avoidant individuals were less likely to achieve remission with EX/RP (odds ratio = 0.04, 95% confidence interval [CI] range 0.01–0.28, P = .001). Baseline avoidance was also associated with degree of patient adherence to between-session EX/RP assignments, which mediated the relationship between baseline avoidance and EX/RP outcomes (P < .05). Baseline avoidance did not predict outcomes or wellness among patients receiving risperidone or placebo. CONCLUSIONS: These results suggest that avoidance behaviors are an important clinical factor in EX/RP outcomes and indicate that assessing avoidance may provide an efficient method for predicting EX/RP outcomes. Avoidance may be particularly relevant in EX/RP as compared to medication treatment, though future replication of these initial results is required.