In order to assess the intrarenal mechanisms responsible for the natriuretic action of caffeine, the renal clearances of (51)Cr-EDTA [used as a measure of glomerular filtration rate (GFR)] and lithium (used as an index of end-proximal fluid delivery) were measured in eight healthy males before (control period) and immediately after (experimental period) a 400 mg oral dose of caffeine (given over 90 min) or placebo. In caffeine-treated subjects, the fractional excretion of sodium rose from 1.00+/-0.25% in the control period to 1.47+/-0.18% in the experimental period, while corresponding values on the placebo day were 1.04+/-0.16% and 0.70+/-0.07% respectively. GFR was unchanged following either caffeine or placebo. When compared with the placebo day, caffeine caused increases in lithium clearance (experimental period values: caffeine, 37+/-1 ml/min; placebo, 28+/-2 ml/min; P <0.001), the fractional excretion of lithium (caffeine, 34+/-1%; placebo, 26+/-2%; P <0.001) and the sodium/lithium clearance ratio (used as an index of the fraction of sodium delivered to the distal nephron that escapes reabsorption therein: caffeine, 4.4+/-0.3%; placebo, 2.8+/-0.2%; P <0.001). These results suggest that reduced fractional sodium reabsorption in both the proximal tubule and the distal nephron contributes to the acute natriuretic effect of caffeine. The data also confirm the importance of controlling caffeine intake when investigating renal function using lithium clearance.