There is a longstanding debate about whether helping welfare recipients quickly find work or helping them to first obtain some basic education and training better improves their economic well-being. This brief contributes to the debate by presenting long-term findings from three sites in the seven-site National Evaluation of Welfare-to-Work Strategies (NEWWS), a multiyear random assignment study designed to directly and very reliably test the effects of alternative approaches to helping welfare recipients find jobs. As part of NEWWS, the research team followed study sample members for five years. This brief extends the follow-up period, presenting the effects of the interventions during a period that falls roughly 10 to 15 years after individuals entered the study. In the first two years of follow-up, a much higher proportion of enrollees in "labor force attachment" (LFA) programs than those in "human capital development" (HCD) programs conducted a job search, and a much higher proportion of enrollees in HCD programs than those in LFA programs participated in education or training, reflecting the two program models' different emphases. People who were randomly assigned to the LFA or HCD groups had more earnings over the entire 15-year follow-up period, on average, than did their control group counterparts, with the largest additional earnings occurring during the first five years after study entry. The LFA approach resulted in substantially higher earnings than did the HCD approach during the first two years of follow-up. In contrast, during follow-up years 10 through 15, enrollees assigned to the HCD group actually earned more than those assigned to the LFA group, although these differences are not statistically significant.