Objective: To describe dietary habits in the Turks and Caicos Islands. Design and Methods: Food frequency questionnaires were administered to female-household-heads of 144 households randomly selected from three islands ' voter's lists (Grand Turk [n = 48], Providenciales [n = 46] and Middle Caicos [n = 50]). Data were collected on the distribution of:(a) Households among Levels 0 -7 of a Food Group Scale, developed using the Cornell Technique of Scaling Dichotomous Data, and based on number of households that consumed seven food groups (meat and legumes, bread/cereals, fruits, vegetables, starchy roots/tubers/fruits; dairy and beverages) weekly; (b) Foods among four categories (common core, island core, occasional or rare) also based on weekly frequency of consumption. Results: Thirty per cent of households on Grand Turk and 37% on Providenciales were at level 7, the most varied and complex diets, compared to 3% for Middle Caicos, which exemplified the indigenous diet of local seafood, beans, and grits (corn) supplemented with imports eg rice and bread/flour. Middle Caicos had substantially fewer island core foods ([n = 16] from four food groups) than did Grand Turk (n = 29) and Providenciales (n = 30), which represented the 7-food groups and included 15 (94%) of Middle Caicos' island core foods. Conclusion: Providenciales and Grand Turk had more varied and complex diets. Understanding how various islands supplement the indigenous/traditional diet is imperative to develop and evaluate (a) island-specific nutrition intervention eg culturally appropriate nutrition education messages (eg to increase iron consumption); and (b) future research protocols.