While physiologic and epidemiologic evidence link diuretic therapy with hyperuricemia, no previous study has quantified the risk for initiation of treatment specific for hyperuricemia or gout among elderly patients taking thiazide diuretics. We performed a retrospective cohort study of 9249 enrollees aged 65 or older in the New Jersey Medicaid program who were newly started on an antihypertensive medication from November 1981 through February 1989 and who had no prior use of anti-gout therapy (allopurinol, colchicine, or a uricosutic) during the preceding one-year period. We used Cox proportional hazards analysis to determine the risk for the initiation of anti-gout therapy in patients using various antihypertensive treatment regimens relative to no antihypertensive exposure. Patient follow-up extended for up to two years. Antihypertensive exposure was characterized over the entire period of follow-up according to the following categories: thiazide diuretic therapy alone; non-thiazide antihypertensive therapy; thiazide diuretic therapy in combination with any non-thiazide antihypertensive agent(s); and no antihypertensive use. Antihypertensive exposure was entered into the model as a time-varying covariate. Estimates of risk were adjusted for age, sex, race, nursing home residence, number of prescriptions filled, intensity of physician use, hospitalization history, and year of antihypertensive treatment initiation. The adjusted relative risk for the initiation of anti-gout therapy was 1.00 (95% CI, 0.65-1.53) for non-thiazide antihypertensive therapy alone, 1.99 (95%, CI, 1.21-3.26) for thiazide diuretic therapy, and 2.29 (95% CI, 1.55-3.37) for thiazide diuretic therapy in combination with any non-thiazide agent(s). Risk for anti-gout therapy was significantly increased for thiazide doses of > or = 25 mg/day (in hydrochlorothiazide equivalents); no significant increase in risk was seen for lower doses. We conclude that use of thiazide diuretics in doses of 25 mg/day or higher is associated with a significantly increased risk for initiation of anti-gout therapy. Such treatment may reflect the occurrence of clinical sequelae of diuretic-induced hyperuricemia or the inappropriate treatment of asymptomatic hyperuricemia.