Objective To test cognitive behavioral therapy (CBT) for persistent attention‐deficit hyperactivity disorder (ADHD) symptoms in a sample of medication‐treated adolescents. Methods Forty‐six adolescents (ages 14–18), with clinically significant ADHD symptoms despite stable medication treatment were randomly assigned to receive CBT for ADHD or wait list control in a cross‐over design. Twenty‐four were randomized to CBT, 22 to wait list, and 15 crossed‐over from wait list to CBT. A blind independent evaluator (IE) rated symptom severity on the ADHD Current Symptom Scale, by adolescent and parent report, and rated each subject using the Clinical Global Impression Severity Scale (CGI), a global measure of distress and impairment. These assessments were performed at baseline, 4‐months (post‐CBT or post wait list), and 8‐months (post‐treatment for those originally assigned to the wait list condition and 4‐month follow‐up for those originally assigned to CBT). Trial Registration: http://clinicaltrials.gov/show/NCT01019252. Results Using all available data, mixed effects modeling, and pooling for the wait list cross‐over, participants who received CBT received a mean score 10.93 lower on the IE‐rated parent assessment of symptom severity (95% CI: −12.93, −8.93; p < .0001), 5.24 lower on the IE‐rated adolescent assessment of symptom severity (95% CI: −7.21, −3.28; p < .0001), and 1.17 lower IE‐rated CGI (95% CI: −1.39, −.94; p < .0001). Results were consistent across 100 multiple imputations (all p < .0001). There was a greater proportion of responders after CBT by parent (50% vs. 18%, p = .00) and adolescent (58% vs. 18% p = .02) report. Conclusions This study demonstrates initial efficacy of CBT for adolescents with ADHD who continued to exhibit persistent symptoms despite medications.