Introduction: Patients often refer to information on the Internet when selecting a medical institution, and some patients provide feedback on their experiences. In this study, we analyzed the content of patients' evaluations of medical institutions on Google.Methods: This study evaluated Google reviews and ratings of medical institutions in Shizuoka Prefecture. We coded the reviews with 12 items according to their content, and further categorized them into "positive," "negative," "unclassifiable," and "no description." We used modified Poisson regression analysis to investigate the relationship between ratings and assessment items.Results: Our sample consisted of 2,044 medical institutions. The number of reviews included in the analysis was 13,769. Reviews frequently commented on "doctor's behaviors," and positive comments about doctor's behaviors were significantly associated with high ratings (B: 0.76, 95%CI: 0.70 to 0.82), whereas negative comments were associated with low ratings (−4.65, −5.24 to −4.06).Conclusion: Within the reviews on Google, doctors' behavior had an impact on the ratings of medical institutions.