Young adults who had previously participated in a longitudinal survey of youth were sent a questionnaire. They were randomly assigned to receive a $20 prepayment, a $20 postpayment, or a $25 postpayment for participation in the latest survey. Those in the large incentive condition were 7 percentage points more likely to return a survey than those in the smaller, postpayment group. Prepayment had a smaller, less reliable effect. Effects of incentive magnitude and timing were consistent at each month of the study period; only better high school grades distinguished early responders from late responders. Nonresponders had characteristics suggestive of low social conformity and were more likely than responders to be African American and male and have low SES. The discussion centers on motivations for participating in research and differences in the incentives likely to promote continued response versus initial study enrollment.