Purpose <p>Anxiety disorders are widespread and significantly impair the mental health and daily functioning of autistic children and adolescents. Internet-delivered cognitive behavioral therapy (iCBT) has emerged as a widespread, low-cost alternative to traditional CBT, helping to reduce demands on therapists without sacrificing efficacy. However, the predictors of treatment outcomes in iCBT for this population remain understudied. This study used data from a randomized clinical trial evaluating parent-led, therapist-supported iCBT among 57 autistic youth with anxiety disorders to examine whether repetitive behaviors and restricted interests, pretreatment anxiety severity, parent-rated externalizing symptoms, verbal intellectual ability, and Hispanic/Latino ethnicity predicted treatment outcomes.</p> Methods <p>Hierarchical linear modeling was used to test the impact of these predictors on changes in evaluator-rated and parent-reported anxiety severity.</p> Results <p>Higher baseline anxiety severity was associated with greater improvement in evaluator- and parent-rated anxiety scores across the trial, while fewer externalizing behavior problems predicted more improvement in parent-reported anxiety and trended towards predicting greater improvement in evaluator-rated anxiety. The other predictor variables did not significantly influence rates of change in anxiety across the trial.</p> Conclusion <p>These findings suggest that youth with varying levels of anxiety can benefit from a light-touch therapeutic approach, challenging the general notion that low-intensity interventions are only viable for individuals with mild anxiety. These findings also suggest that additional considerations are needed for youth with more severe disruptive behavior problems. Replication of these findings in larger samples with comparison groups (e.g., an inert comparator or a comparison to standard intensity treatment) is highlighted for future work.</p>