ABSTRACT: BACKGROUND:African Americans suffer more than non-Hispanic whites from type 2 diabetes, but diabetes self-management education (DSME) has been less effective at improving glycemic control for African Americans. Our objective was to determine whether a novel, culturally tailored DSME intervention would result in sustained improvements in glycemic control in low-income African-American patients of public hospital clinics. RESEARCH DESIGN AND METHODS:This randomized controlled trial (n =?211) compared changes in hemoglobin A1c (A1c) at 6, 12, and 18 months between two arms: (1) Lifestyle Improvement through Food and Exercise (LIFE), a culturally tailored, 28-session community-based intervention, focused on diet and physical activity, and (2) a standard of care comparison group receiving two group DSME classes. Cluster-adjusted ANCOVA modeling was used to assess A1c changes from baseline to 6, 12, and 18 months, respectively, between arms. RESULTS:At 6 months, A1c decreased significantly more in the intervention group than the control group (-?0.76 vs -?0.21%, p =?0.03). However, by 12 and 18 months, the difference was no longer significant (12 months -?0.63 intervention vs -?0.45 control, p =?0.52). There was a decrease in A1c over 18 months in both the intervention (? =?-?0.026, p =?0.003) and the comparison arm (? =?-?0.018, p =?0.048) but no difference in trend (p =?0.472) between arms. The intervention group had greater improvements in nutrition knowledge (11.1 vs 6.0 point change, p =?0.002) and diet quality (4.0 vs -?0.5 point change, p =?0.018) while the comparison group had more participants with improved medication adherence (24% vs 10%, p