Mailed questionnaires are an economical method of data collection for epidemiologic studies, but response tends to be lower than for telephone or personal interviews. As part of a follow-up study of volunteers who provided a brief health history and blood sample for a blood specimen bank in 1989, the authors conducted a controlled trial of the effect of length, incentives, and follow-up techniques on response to a mailed questionnaire. Interventions tested included variations on length of the questionnaire, effect of a monetary incentive, and effect of a postcard reminder versus a letter accompanied by a second questionnaire. Response was similar for the short (16-item, 4-page) and long (76-item, 16-page) questionnaire groups. The non-monetary [corrected] incentive did not improve the frequency of response. The second mailing of a questionnaire was significantly better than a postcard reminder in improving responses (23% vs. 10%). It is important to systematically test marketing principles to determine which techniques are effective in increasing response to mailed questionnaires for epidemiologic studies.