OBJECTIVE: To determine if Burmese cats in Queensland have an increased risk of diabetes mellitus. DESIGN: A retrospective study of diabetic and nondiabetic cats that had blood submitted to a veterinary clinical laboratory over a 22 month study period. SAMPLE POPULATION: 4402 cats PROCEDURE: Cats were considered diabetic if blood glucose concentration was > 11 mmol/L and fructosamine was > 406 mumol/L or hydroxybutyrate was > 1 mmol/L. Cats were grouped into Burmese and non-Burmese. Adjusted odds ratios of diabetes were calculated for breed, gender and age group amongst cats with blood glucose > 11 mmol/L. RESULTS: Burmese cats comprised 20% of 45 diabetic cats of known breed, which was higher (P < 0.001) than among the normoglycemic reference population of 2203 cats (7% Burmese). There were more females among the diabetic Burmese (62%), but this did not differ (P > 0.05) from the Burmese reference population (45% females). In contrast, males seemed to predominate among diabetic non-Burmese (63%), although this also did not differ (P > 0.05) from the reference population (55%) of from diabetic Burmese (38% males). The majority (90%) of diabetic cats were older than 6 years, irrespective of breed (median age 12 years, interquartile range 10 to 13 years). This was higher (chi(2) = 8.13, P < 0.005) than among the normoglycaemic reference population, where 69% were older than 6 years. CONCLUSIONS: Burmese cats were significantly over represented among cats with diabetes mellitus. Irrespective of breed, the risk of diabetes in the study population increased with age.