To differentiate between extrarenal and renal causes of hyperuricaemia and gout, clearances of urate and creatinine were monitored for 3 1/2 days in fifty-two individuals (seven with a history of gout) with no gross impairment of renal function (creatinine clearance 52-137 ml/min). Dietary purine intake was kept constant. Monophasic circadian fluctuations of fractional urate excretion (= urate clearance over creatinine clearance) were observed with peak values in the afternoon, about 50% higher than during the night. Circadian fluctuations of urinary flow rate were almost identical. However, enhancement of urinary flow rate due to water diuresis had no effect on urate clearance. Despite wide variation of plasma urate concentrations among different individuals (+/- 30% SD), daily urate excretion varied little (+/- 4% SD) and did not correlate with plasma urate (r = 0.03). Thus extrarenal factors appear not to account for the occurrence of hyperuricaemia in these patients. In contrast, a clearcut negative correlation was apparent between plasma urate concentration and fractional urate clearance (r = 0.72), which could fully account for the variations of plasma urate concentration. To elucidate further the mechanism responsible for antiuricosuria in hyperuricaemic patients, the effects of the uricosuric agents benzbromarone and probenecid were tested. A clearcut correlation was apparent between control fractional urate excretion and uricosuric effect of both benzbromarone and probenecid (r = 0.83 and 0.88, respectively), suggesting that anti-uricosuria was due to defective secretion. In an additional series, the uricosuric effect of probenecid was tested in ten patients with renal insufficiency. In these patients the uricosuric effect was clearly blunted, indicating that urate reabsorption is reduced in renal insufficiency.