After an oral or intravenous dose of furosemide, there is considerable interindividual variability in the amount of unchanged drug delivered into the urine. On average, approximately half as much reaches the intraluminal site of action with an oral compared to an intravenous dose. However, the natriuretic response to the same dose administered by either route is virtually the same. Similarly, after pretreatment with probenecid, the same total amount of furosemide in urine causes a greater overall response. It has been presumed that this paradox is accounted for by differences in rate of delivery of furosemide to the active site such that after an oral dose, or after pretreatment with probenecid, amounts of drug are for longer periods of time at the "steep" portion of the dose-response curve. Our analysis shows this not to be the case. For furosemide, the "slope factor" of the dose-response curve is such that the amount of diuretic delivered into the urine which is maximally efficient (21.5 micrograms/min) is considerably less than the amount causing half-maximal response (69.8 micrograms/min). Oral administration or pretreatment with probenecid maintains drug close to this maximally efficient amount more persistently than does intravenous administration. By so doing, total response to an oral dose approaches that of intravenous dosing despite delivering half the amount of drug to the active site, and after probenecid an intravenous dose causes a greater response than intravenous dosing alone despite delivering the same amount of drug to the active site. These data emphasize the importance of the time course of delivery of drug to the active site as an independent determinant of overall response.