Over the 4-year study, a high retention rate of 91 percent was achieved, and a retention rate of 94 percent was maintained for the first year of a 4-year extension of the original study. The $20 bonus for full-family participation was an incentive for obtaining the participation of the entire family. Participating youth responded well to the small cash reward in addition to their share of the family payment. Participants of all ages anticipated bigger payments with each subsequent year and hoped to win the yearly drawings. The second most influential element of the retention package was the use of flexible, convenient, and home-based assessments. By scheduling home visits when the family was available ensured that assessments were convenient for the family. The third important factor in retention involved trust. Providing each family with the same assessor throughout the study gave each family a personal connection with the research organization. This was especially true for African-American families. The African-American assessor was involved in the area's small but close-knit African-American community. Of the variables of age, ethnicity, income, sex, parent/guardian marital status, parent alcohol use, and family cohesion, only sex significantly predicted retention; girls were more likely than boys to remain in the study. Ethnicity nearly reached significance. The study involved collecting data annually for 4 years from residents of a large metropolitan area in the Pacific Northwest. Of the sample of 405 youth and their families, the youth were 48.4 percent female, 50.4 percent African-American, and 49.6 percent White. 3 tables and 26 references