In a longitudinal, multicenter study of 4,954 men at risk for human immunodeficiency virus infection and acquired immunodeficiency syndrome, data from the first 9.5 years of follow-up (April 1984 through Septembe 1993) were used to determine differences between those who remained in the study and those who dropped out. Demographic variables (age, race, education, employment, and study center), health status (human immunodeficiency virus type 1 serostatus and depression), and behavioral characteristics (alcohol drinking, drug use, and anal-receptive intercourse) were analyzed. Strategies for promoting retention included having frequent contact with participants, generating trust, keeping participants well-informed, utilizing multiple resources for follow-up, and providing flexible methods of participation. After 9.5 years of follow-up, vital status was known for 4,385 (88.5%) of the participants. Results from multiple logistic regression showed that race, age, education, and smoking were each significantly associated with nonparticipation (p < 0.001). A high level of retention was maintained in this well-educated and highly motivated cohort of homosexual/bisexual men. Extensive follow-up methods may improve case-finding. Nonwhite race, younger age, less education, and smoking were important predictors of dropping out. These findings identify specific groups for targeting follow-up efforts to reduce potential bias due to dropout.