Thiazide diuretics are considered to be the cornerstone of contemporary antihypertensive therapy and are generally recommended as the initial treatment for patients with mild to moderate, uncomplicated hypertension. Hypokalemia and hypomagnesemia are two metabolic alterations that are associated with long-term thiazide therapy. Thirty-five patients (20 with low renin status and 15 with normal renin status) with mild to moderate essential hypertension were treated with hydrochlorothiazide in a dose-titration experiment after a four-week lead-in period. The initial daily dose was 12.5 mg; this was increased at four-week intervals to 25 mg, 37.5 mg, and 50 mg daily. The endpoint dose of this titration was that dose at which the patient's blood pressure normalized, or the dose of 50 mg, if that dose was reached. Patients were maintained on their endpoint dose of hydrochlorothiazide for 24 weeks of continuous thiazide monotherapy beyond the dose titration. The serum potassium and serum magnesium levels during the control period were 4.4 +/- 0.2 mmol/liter and 2.30 +/- 0.08 mg/dl, respectively. During dose titration, each incremental increase of hydrochlorothiazide produced a decrease in blood pressure and a stepwise decrease in serum potassium and magnesium levels. A previously reported study involving 38 patients with mild to moderate hypertension (22 with low renin status and 16 with normal renin status) used similar methods to study higher-dose thiazide therapy. An initial dose of 50 mg daily of hydrochlorothiazide was administered; this was increased at four-week intervals to 100 mg, 150 mg, and 200 mg. The serum potassium and serum magnesium levels during the control period were 4.5 +/- 0.2 mmol/liter and 2.1 +/- 0.18 mg/dl, respectively. In the hypertensive patients with normal renin status, doses of hydrochlorothiazide greater than 50 mg did not result in further blood pressure lowering effects; however, the undesirable effects of hypokalemia and hypomagnesemia continued to be manifested and increased at higher doses of hydrochlorothiazide. Thirty-eight patients who had previously experienced hypokalemia, palpitations, or cardiac arrhythmia were placed on hydroclorothiazide therapy for one to three months and were monitored for arrhythmias after treadmill exercise. The occurrence of premature ventricular contractions correlated significantly with the decrease in serum potassium (r = 0.73, p less than 0.001) and serum magnesium (r = 0.68, p less than 0.001) levels during hydrochlorothiazide therapy and with the product of the change of the two cations (r = 0.81, p less than 0.001).