Thirty-three patients with hypertension receiving drug treatment that included a potassium losing diuretic were randomly allocated to 64 mmol of potassium (14 patients) or to no additional potassium supplementation (19 patients). Potassium was administered as slow release potassium chloride. After 3 months, blood pressure fell by 5/1 mm Hg in the patients who received the supplements and by 7/4 mm Hg in those who did not receive them. The falls in pressure were not significantly different and the 90% confidence limits for the effect of supplementation on diastolic pressure were: a fall of 2 mm Hg and a rise of 11 mm Hg, thus excluding an important hypotensive effect in these patients. Conversely, plasma creatinine fell by 11% in the supplement group compared with a 6% rise in the control group (P less than 0.05). Potassium supplementation, either by pharmacological preparations or by dietary manipulation, may prove to be desirable in patients on a potassium losing diuretic but should not be expected to lower blood pressure in such patients.