Homogeneous groups surveyed about issues of concern to them may require few follow-ups because respondents provide role-specific answers representative of others in that role. The authors analyzed three questionnaire mailings to 1,535 physicians that produced 977 responses (604 first, 252 second, and 85 third mailing). The only demographic or substantive variable significantly different between mailings was medical specialty. This supports the hypothesis that surveys of homogeneous groups, such as physicians in one specialty, may require fewer follow-up efforts. However, using few follow-ups must be done cautiously to avoid nonresponse bias and insure sufficient sample size.The role of physician involves membership in a relatively homogej neous professional group. Physicians are both self-selected and screened by admissions committees, and also subjected to a long and standardized process of socialization into the physician role. As important decision makers in the health care system, physicians are frequently surveyed AUTHORS' NOTE: The authors thank Susan Malamet for assistance tn selecting the sample.