The present article uses meta-analysis to examine treatment components related to outcome within 35 randomized controlled trials (RCTs) for childhood anxiety disorders (CADs) and eight RCTs for childhood obsessive compulsive disorder (OCD). Examination of the RCTs of cognitive behavioral therapy (CBT) for CADs suggested that adding relaxation and delaying exposures until after the introduction of other anxiety management strategies does not increase the efficacy of exposure-based treatment. In addition, compared to the large effect size (ES) associated with exposure and response prevention (ERP) for OCD (k = 9, mean ES = 1.93), the effect size associated with CBT for CADs (k = 44, mean ES = 0.89) did not differentiate from attention placebo (k = 11, mean ES = 0.55), although it was more effective than waitlist control (k = 24, mean ES = 0.22). Instructively, ERP for OCD involved more exposure initiated earlier and less relaxation than CBT for CADs. In addition, RCTs of ERP were more likely to use clinician-administered measures as opposed to self-report and to be conducted in clinical versus recruited samples. These results suggest that dismantling studies using a gold-standard clinician-rated outcome measure to compare the value of adding anxiety management strategies to exposure will be necessary to increase the efficacy of CBT for CADs to levels achieved by ERP for OCD.