Six patients with chronic renal failure (CRF group) and four healthy subjects (HS group) were given 5 mg oral and intravenous doses of bumetanide in a random, crossover design. The CRF group had significantly lower plasma and renal clearances, resulting in a five-to sixfold reduction in the fractional urinary excretion of the drug. The percent free drug in plasma for the CRF group was more than double that for the HS group, and significant correlations were observed for volume of distribution at steady state vs. percent free (r = 0.661; P less than 0.05), nonrenal clearance vs. percent free (r = 0.796; P less than 0.01), and renal clearance vs. creatinine clearance (r = 0.995; P less than 0.001). Although bioavailability was relatively consistent among the HS (0.664 +/- 0.112) and CRF (0.689 +/- 0.149) groups, the absorption-time profiles were more irregular for both groups. Cumulative sodium excretion and overall efficiency of response to bumetanide did not differ significantly between the two routes of administration in either group.