IMPORTANCE: Clinician-guided online self-help based on cognitive behavioral therapy (CBT) has been shown to be effective at decreasing symptom severity for people with atopic dermatitis (AD). A brief online self-guided CBT intervention could be more cost-effective and allow for easy implementation and broader outreach compared with more comprehensive clinician-guided interventions. OBJECTIVE: To investigate whether a brief online self-guided CBT intervention is noninferior to a comprehensive online clinician-guided CBT treatment. DESIGN, SETTING, AND PARTICIPANTS: This single-blind randomized clinical noninferiority trial was conducted at Karolinska Institutet, Stockholm, Sweden. Adult individuals with AD were enrolled from November 2022 to April 2023. The last postintervention data were collected in December 2023. INTERVENTIONS: Participants randomized to the self-guided group had access to a self-guided online CBT intervention for 12 weeks without clinician support. Participants randomized to the clinician-guided group received online CBT for 12 weeks. MAIN OUTCOMES AND MEASURES: The primary outcome was change in score from baseline to postintervention to 12-week follow-up on the self-reported Patient-Oriented Eczema Measure (POEM). The predefined noninferiority margin was 3 points on POEM. RESULTS: Of 168 randomized participants, 142 (84.5%) were female, and the mean (SD) age was 39 (10.5) years. A total of 86 participants were randomized to the self-guided group and 82 were randomized to the clinician-guided group. A total of 151 (90.0%) completed the main outcome postintervention assessment. Postintervention, the clinician-guided group had improved 4.20 points (95% CI, 1.94-6.05) on POEM and the self-guided group improved 4.60 points (95% CI, 2.57-6.64), corresponding to an estimated mean difference in change of 0.36 points (1-sided 97.5% CI, -∞ to 1.75), which was below the noninferiority margin of 3 points. No serious adverse events were reported. In the clinician-guided group, clinicians spent a mean (SD) of 36.0 (33.3) minutes (95% CI, 29.2-41.7) on treatment guidance and 14.0 (6.0) minutes (95% CI, 12.9-15.6) on assessments compared to 15.8 (6.4) minutes on assessments in the self-guided group. CONCLUSIONS AND RELEVANCE: In this randomized clinical noninferiority trial, a brief self-guided CBT intervention was noninferior to clinician-guided CBT. Given the limited clinical resources required to deliver self-guided CBT, this treatment might be a promising means to disseminate evidence-based psychological treatment for patients with AD. TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT05517850.