Background: Exposure tasks—where an individual confronts a feared stimulus or situation—are known to be a key element of the treatment for youth anxiety. However, optimal therapist behaviors during these exposure tasks and the specifics of how therapist should conduct exposure tasks have not been determined. The current study examined the relationship between therapist behaviors that (a) increased, (b) decreased or (c) maintained the youth’s anxiety during exposures and treatment outcomes. Methods: Participants were youth (N = 107) ages 7 to 17 who received cognitive behavioral therapy for anxiety. Youth and their primary caregiver(s) completed a diagnostic interview and self- and parent-report measures pre- and post-treatment. Exposure session videos were rated by observers trained to reliability on a coding system evaluating therapist behaviors. Hierarchical regression analyses examined the role of therapist behaviors in predicting treatment outcomes. Logistic regression assessed the ability of therapist behaviors to predict treatment responder status (i.e., being a treatment responder versus a non-responder). Exploratory analyses examined the relationship between the individual therapist behaviors (within the three overall categories of behaviors) and treatment outcomes. Results: Youth showed significant improvement over the course of treatment. The three categories of therapist behaviors used during exposure tasks (increase, decrease and maintain the youth’s anxiety) were not associated with treatment outcomes. Discussion: Findings indicate that so long as exposure tasks are conducted, the therapist behaviors during the exposures may not be as important for predicting outcomes. Clinical implications, study limitations, and future directions are discussed. / Psychology