PURPOSE: Patients have been increasingly using physician-rating websites (PRWs); however, few studies have analyzed the validity of star ratings on PRWs. We aimed to compare PRW patient satisfaction scores with internally generated patient satisfaction scores (internal scores) of physicians at a large quaternary cancer center. METHODS: We collected internal scores and PRW scores for physicians at MD Anderson Cancer Center. Internal scores were based on patient responses to the Clinician and Group Consumer Assessment of Healthcare Providers and Systems patient experience (CG-CAHPS) survey. Only physicians with an internal score on the basis of ≥ 30 patient reviews were included. The median numbers of reviews and median scores were compared between internal data and four PRWs (Google, HealthGrades, Vitals, and WebMD). Both internally and on PRWs, possible scores ranged from 1 (least satisfied) to 5 (most satisfied). RESULTS: Of 640 physicians with an internal score, 510 (79.7%) met the inclusion criteria. For these 510 physicians, the median (IQR) number of internal reviews was 49.5 (30-93) and the median (IQR) internal score was 4.89 (4.81-4.93); the median number of reviews on PRWs ranged from 2 to 7, and the median score on PRWs ranged from 4.40 to 5.00. No physician had an internal score < 4, but the proportions with score < 4 on PRWs ranged from 16% to 30%. CONCLUSION: Internal patient satisfaction scores were higher and calculated from more reviews than PRW patient satisfaction scores and correlated weakly with PRW scores. Given that patients rely on PRWs when evaluating potential physicians, we recommend publishing internal scores online to give patients more complete information regarding physician performance.