This study investigates young adults’ first experience with the Supplementary Nutrition Assistance Program (SNAP), examining the determinants of first program entry and exit. It makes use of the National Longitudinal Survey of Youth 1997 cohort (NLSY97), which follows respondents from adolescence into adulthood. This study estimates discrete-time hazard models of program entry and exit with and without unobserved heterogeneity. Unobserved heterogeneity is modeled using both a parametric approach, in which a gamma distribution is assumed, and a non-parametric approach with two mass points. The results are broadly consistent across models, indicating that, for the cohort in this study, accounting for unobserved heterogeneity does not substantially alter the results from a basic discrete-time hazard model. The results show that expanded categorical eligibility increased the hazard of SNAP entry in the six years following high school, while the absence of vehicle exclusions decreased the entry hazard. For program exit, however, state SNAP policies had no statistically significant effect. The recent birth of a child, prior participation in WIC and low educational attainment were each strongly associated with an increased “risk” of SNAP entry, and decreased “risk” of exit. Somewhat, surprisingly, higher unemployment rates in the local labor market were not significantly associated with higher entry risk, but were strongly associated with a lower exit risk.