Measures of association in prospective studies can be distorted by incomplete follow-up. Various mailing strategies were used to contact 12,233 cohort members of the Health Professionals Follow-up Study who had not responded to three successive bulk-rate mailings. Response rates were highest, 79.5% overall, from participants who were sent a certified mailing in phase 1 (63.2%), followed by a repeat certified mailing to nonrespondents (44.3%). Although altering the physical appearance of the envelope and using other postal rates were tested, certified mail was the most effective approach for reaching study members who were nonrespondents to a mailed questionnaire.