Epidemiologic studies suggest an inverse relation between potassium intake and the prevalence of hypertension. To investigate the effect of dietary potassium restriction on blood pressure, we used a randomized crossover design to study 10 healthy, normotensive men randomly assigned to isocaloric diets (each lasting nine days) providing either low (10 mmol per day) or normal (90 mmol per day) amounts of potassium, while sodium intake was maintained at the subjects' usual levels (120 to 200 mmol per day). With the low-potassium diet, plasma potassium levels declined from 3.8 to 3.2 mmol per liter (P less than 0.001), but plasma sodium and chloride levels were unchanged. The average daily excretion of urinary sodium (+/- SEM) on the low-potassium diet was significantly lower than that with the normal-potassium diet (10 +/- 10 vs. 144 +/- 10 mmol; P less than 0.001). The mean arterial pressure did not change significantly during normal potassium intake, but it increased over the nine days of the low-potassium diet from 90.9 +/- 2.2 to 95.0 +/- 2.2 mm Hg (P less than 0.05). Both mean arterial (P less than 0.01) and diastolic (P less than 0.005) pressures were significantly higher after the low-potassium diet than after the normal-potassium diet. Potassium depletion suppressed plasma aldosterone levels but had no effect on plasma renin activity or on arginine vasopressin or catecholamine levels. A saline infusion further increased the mean arterial pressure in the potassium-depleted subjects but had no effect in the control group (P less than 0.05). We conclude that short-term potassium depletion increases blood pressure in healthy, normotensive men and permits further increases in blood pressure after saline loading. We found no evidence that the hypertensive effect of potassium depletion resulted from changes in either renal hemodynamics or circulating levels of vasoactive hormones.