OBJECTIVE Marshallese adults experience high rates of type 2 diabetes. Previous diabetes self-management education (DSME) interventions among Marshallese were unsuccessful. This study compared the extent to which two DSME interventions improved glycemic control, measured on the basis of change in glycated hemoglobin (HbA1c). RESEARCH DESIGN AND METHODS A two-arm randomized controlled trial compared a standard-model DSME (standard DSME) with a culturally adapted family-model DSME (adapted DSME). Marshallese adults with type 2 diabetes (n = 221) received either standard DSME in a community setting (n = 111) or adapted DSME in a home setting (n = 110). Outcome measures were assessed at baseline, immediately after the intervention, and at 6 and 12 months after the intervention and were examined with adjusted linear mixed-effects regression models. RESULTS Participants in the adapted DSME arm showed significantly greater declines in mean HbA1c immediately (20.61% [95% CI 21.19, 20.03]; P = 0.038) and 12 months (20.77% [95% CI 21.38, 20.17]; P = 0.013) after the intervention than those in the standard DSME arm. Within the adapted DSME arm, participants had significant reductions in mean HbA1c from baseline to immediately after the intervention (21.18% [95% CI 21.55, 20.81]), to 6 months (20.67% [95% CI 21.06, 20.28]), and to 12 months (20.87% [95% CI 21.28, 20.46]) (P < 0.001 for all). Participants in the standard DSME arm had significant reductions in mean HbA1c from baseline to immediately after the intervention (20.55% [95% CI 20.93, 20.17]; P = 0.005). CONCLUSIONS Participants receiving the adapted DSME showed significantly greater reductions in mean HbA1c immediately after and 12 months after the intervention than the reductions among those receiving standard DSME. This study adds to the body of research that shows the potential effectiveness of culturally adapted DSME that includes participants’ family members.