The decline in first-phase insulin secretion is a key event in the etiology of type 2 diabetes (T2D). Although the cause of beta-cell failure is not clear, “lipotoxicity” has been proposed. Bariatric surgery and very-low calorie diets in patients with T2D induce disease remission, characterized by a return of first-phase insulin secretion and a depletion of pancreas lipid. However, these are extreme approaches to treating T2D, and non-invasive, sustainable, yet equally effective, treatments are needed. We have shown in individuals at risk for T2D that an intervention with a weight-maintaining low-glycemic (LG) diet selectively depletes visceral adipose tissue and ectopic lipid in muscle while preserving thigh subcutaneous adipose and lean body mass. This observation suggests that such diets are able to “remodel” body composition by re-partitioning energy away from metabolically harmful lipid stores. Participants on the LG diet also demonstrated improved insulin sensitivity and a dramatic (9-fold) increase in first-phase insulin secretion. Thus, we hypothesize that a weight-maintaining LG diet will selectively deplete ectopic adipose tissue, including pancreatic lipid, and will permit recovery of beta-cell function in individuals with T2D. Rescue of beta-cell function may be particularly important in African-Americans (AA), who as a group demonstrate a high prevalence of T2D, for reasons that cannot be explained by lifestyle. AA are likely to be vulnerable to beta-cell failure due to inherently high beta-cell responsiveness (demonstrable in healthy young children). Further, it has been shown that pancreas lipid is a determinant of prediabetes specifically in AA. Thus, we hypothesize that an LG diet will be particularly beneficial to beta-cell function and glycemic control among AA. To test these hypotheses, we will conduct a randomized clinical trial to examine the impact a weight-maintaining LG diet vs a Control diet (ADA/USDA) with all food provided on changes in pancreatic lipid by magnetic resonance imaging (MRI) and first-phase insulin secretion by hyperglycemic clamp and oral meal test in AA and European-American (EA) individuals with early T2D (<5 yr since diagnosis). The study will be powered to examine race-specific effects (race x diet interaction). This proposal responds to PA-17-021, “Addressing Health Disparities in NIDDK Diseases.” The results from this study could change clinical care guidelines to incorporate an LG diet, which may reduce the disproportionate burden of T2D and its complications experienced by AA.