The purpose of this study was to investigate the association between 25-hydroxyvitamin D [25(OH)D] levels and diabetes status in three ethnicities. This cross sectional study included Cuban Americans (n = 199), Haitian Americans (n = 253) and African Americans (n = 248) with and without type 2 diabetes (T2D) from Florida, U.S. Recruitment of participants was through mailing lists (Cuban Americans and African Americans) and community based sources (Haitian Americans). Adjusted logistic regression analysis indicated significant differences in the odds of having insufficient 25(OH)D levels by ethnicity [P < 0.001], diabetes status [P < 0.001], and their interaction [P < 0.001]. Holm's modified Bonferroni method showed that only Cuban Americans without T2D had significantly lower odds of having insufficient 25(OH)D compared to all other groups. Haitian American and African American participants and participants having T2D had the greatest risk of 25(OH)D insufficiency. Future studies should focus on the role of vitamin D supplementation and diabetes outcomes across ethnic groups.