Increased metabolism of pindolol in renal impairment has previously been suggested by pharmacokinetic calculations. The present study was a pharmacokinetic and metabolic investigation in 7 patients with severe renal impairment (endogeneous creatinine clearance below 5 ml/min). All the patients received pindolol 5 mg t.d.s. 5 days. On the sixth day, after an overnight fast, 14C-pindolol 5 mg was given orally as a solution to drink. Blood samples were taken for up to 72 h and urine was collected at intervals up to 96 h for measurement of unchanged pindolol by a fluorimetric method and total radioactivity by liquid scintillation counting. Metabolites in blood and urine were analysed after separation by HPLC. It was found that the plasma levels following a single dose of 14C-pindolol were similar to those observed in healthy volunteers, but the elimination half-life was slightly increased u tp 11.5 h. The observed steady state plasma concentrations of pindolol were twice as high but they are still in the therapeutic range of 10 to 100 ng/ml. Therefore, the dose of pindolol could have been reduced by a factor 2, but the reduction was not essential. No active metabolite of pindolol was found in plasma or urine, but elimination of the metabolites was decreased. The elimination half-life following multiple doses was prolonged compared to normal and it was quite comparable to that found fort pharmacodynamic half-life in renal patients. The discrepancy between the present findings and the previous results for metabolism and pharmacodynamic half-life was probably due to the sensitivity of the fluorimetric assay of pindolol.