IMPORTANCE Mohs micrographic surgery (MMS) is a skin cancer treatment that uses staged excisions based on margin status. Wide surgeon-level variation exists in the mean number of staged resections used to treat a tumor, resulting in a cost disparity and question of appropriateness. OBJECTIVE To evaluate the effectiveness of a behavioral intervention aimed at reducing extreme overuse in MMS, as defined by the specialty society, by confidentially sharing stages-per-case performance data with individual surgeons benchmarked to their peers nationally. DESIGN, SETTING, AND PARTICIPANTS This nonrandomized controlled intervention study included 2329 US surgeons who performed MMS procedures from January 1, 2016, to March 31, 2018. Physicians were identified using a 100% capture of Medicare Part B claims. The intervention group included physicians affiliated with the American College of Mohs Surgery, and the control group included physicians not affiliated with the American College of Mohs Surgery. INTERVENTIONS Individualized performance reports were delivered to all outlier surgeons, defined by the specialty society as those with mean stages per case 2 SDs above the mean, and inlier surgeons in the intervention group. MAIN OUTCOMES AND MEASURES The primary outcome was surgeon-level change in mean stages per case between the prenotification (January 2016 to January 2017) and postnotification (March 2017 to March 2018) periods. A multivariable linear regression model was used to evaluate the association of notification with this surgeon-level outcome. The surgeon-level metric of mean stages per case was not risk adjusted. The mean Medicare cost savings associated with changes in practice patterns were calculated. RESULTS Of the 2329 included surgeons, 1643 (70.5%) were male and 2120 (91.0%) practiced in metropolitan areas. In the intervention group (n = 1045), 53 surgeons (5.1%) were outliers; in the control group (n = 1284), 87 surgeons (6.8%) were outliers. Among the outliers in the intervention group, 44 (83%) demonstrated a reduction in mean stages per case compared with 60 outliers in the control group (69%; difference, 14%; 95% CI of difference, −1 to 27; P = .07). There was a mean stages-per-case reduction of 12.6% among outliers in the intervention group compared with 9.0% among outliers in the control group, and outliers in the intervention group had an adjusted postintervention differential decrease of 0.14 stages per case (95% CI, −0.19 to −0.09; P = .002). The total administrative cost of the intervention program was $150 000, and the estimated reduction in Medicare spending was $11.1 million. CONCLUSIONS AND RELEVANCE Sharing personalized practice pattern data with physicians benchmarked to their peers can reduce overuse of MMS among outlier physicians.