The Supplemental Nutrition Assistance Program (SNAP) participants have been shown to have worse dietary behaviors compared to eligible non-participants. It is suggested that individuals who self-select into SNAP have unhealthy diets prior to program participation. This study examined the association between low-income adults’ SNAP participation status and dietary behaviors over time. A sample of households from low-income high-minority cities were surveyed at baseline (T1) and 2-4 years later (T2). Analyses were restricted to adults with household income < 200% of the federal poverty line (n=338). Participants were grouped into the four following categories: Long-term (n=171): SNAP participation at T1 and T2; Previous (n=47): SNAP participation at T1 but not T2; Current (n=38): SNAP participation at T2 but not T1; and Never (n=132): no SNAP participation at T1 or T2. Self-reported daily consumption frequency of whole fruits, fruit juice, vegetables, sugar-sweetened beverages (SSBs), and snacks were measured through self-reports. To observe dietary differences based on SNAP participation status, T2 behaviors of all groups were compared to those of the Never group. To observe dietary differences prior to SNAP participation, T1 behaviors were compared between Current and Never groups. There were no significant differences in dietary behaviors at T1 (prior to SNAP participation) between Never and Current participants providing no evidence of self-selection of unhealthy households into SNAP. Only marginal differences were found between SNAP participation status and dietary behaviors at T2. Improvements in SNAP participants’ diets may benefit from focusing on policy changes that encourage healthy dietary habits during participation in the program.