Single-injection renal clearance methods based on plasma clearance alone, without urine collection, are sometimes met with skepticism. They require data extrapolation to infinite time, which is hard to justify a priori. It has been asserted that they are less accurate for rapidly cleared tubular agents than for slowly cleared glomerular filtration rate agents. In this study, we compare urine-based and urine-free methods for the tubular agents 99mTc-MAG3 and 131I-OIH. METHODS In 18 patients, dual-tracer plasma data were obtained from 4 to 90 min after injection (nine samples). Urine was also collected for 90 min (in two voidings). The urine counts wre corrected for residual bladder activity by pre- and postvoid dual-channel gamma camera images. RESULTS When comparing the two methods of clearance calculations, the difference between urine-based and urine-free measurements 1 +/- 5 ml/min for 99mTc-MAG3 and 23 +/- 8 for 131I-OIH (mean +/- s.e. of the mean). For 99mTc-MAG3, the regression line did not differ significantly from the line of identity. The correlation coefficient was 0.94 for both agents. CONCLUSIONS Urine collection is not necessary to measure renal clearance, even for the rapidly cleared tubular agents, except at low clearance levels (when the small absolute error corresponds to a large percentage error).