The medical group survey from the CAHPS (formerly Consumer Assessment of Health Plans Study) project, G-CAHPS, focuses on patient experiences in receiving care from their medical group practice. We compared mail and telephone responses to the G-CAHPS survey in a sample of 880 patients from four physician groups. Patients were randomly assigned to mode. Analyses included comparison of response rates, missing data, internal consistency reliability of six multi-item scales, and mean scores. A total of 537 phone completes and 343 mail completes were obtained (54% response rate). There were no significant differences in internal consistency by mode. In addition, there was only one significant mode difference in item and composite means by mode of administration after adjusting for case-mix differences. This study indicates that mail and telephone modes of data collection for the G-CAHPS survey produce similar results.