The natriuretic effect of the new loop diuretic piretanide was investigated in patients with severe renal insufficiency and was compared with that of furosemide. In the first study 4 hospitalized patients (serum creatinine 407 to 1220 mumol/l) were examined after administration of piretanide (12, 24, 48 and 96 mg to two patients, and 24, 48, 96 and 192 mg to 2 other subjects, given every third day). In the second study 6 hospitalized patients (serum creatinine 194 to 698 mumol/l) were studied after receiving orally 2 different doses of piretanide and 2 different doses of furosemide orally, given every fourth day. The mean natriuretic effect of 48 mg and 96 mg piretanide was 250 and 340% of the control value for the entire group, and 311 to 480% in the subgroup of patients with serum creatinine below 530 mumol/l. For a given dose the natriuresis was inversely correlated with renal function, and at a given serum creatinine level the natriuretic response was dose-dependent. The drug had less effect on water and potassium diuresis than on natriuresis. No significant difference in natriuretic effect was found on comparison with furosemide given in the ratio furosemide:piretanide 3.33:1. The pharmacokinetic data showed a direct correlation between the dose and the mean plasma concentration and also between urinary recovery of the drug and the measured natriuretic response.