Inorganic arsenic exposure in drinking water has been recently related to diabetes mellitus. To evaluate this relationship the authors conducted in 2003, a case-control study in an arseniasis-endemic region from Coahuila, a northern state of Mexico with a high incidence of diabetes. The present analysis includes 200 cases and 200 controls. Cases were obtained from a previous cross-sectional study conducted in that region. Diagnosis of diabetes was established following the American Diabetes Association criteria, with two fasting glucose values > or = 126 mg/100 ml (> or = 7.0 mmol/l) or a history of diabetes treated with insulin or oral hypoglycemic agents. The next subject studied, subsequent to the identification of a case in the cross-sectional study was taken as control. Inorganic arsenic exposure was measured through total arsenic concentrations in urine, measured by hydride-generation atomic absorption spectrophotometry. Subjects with intermediate total arsenic concentration in urine (63.5-104 microg/g creatinine) had two-fold higher risk of having diabetes (odds ratio=2.16; 95% confidence interval: 1.23, 3.79), but the risk was almost three times greater in subjects with higher concentrations of total arsenic in urine (odds ratio=2.84; 95% confidence interval: 1.64, 4.92). This data provides additional evidence that inorganic arsenic exposure may be diabetogenic.