Comparing the measurements of both glomerular filtration (GFR) and tubular excretion rates [TER(MAG3)] by multi-sample and single-sample methods has been performed after a single bolus injection of 3.7 MBq 51Cr-EDTA plus 37 MBq 99mTc-MAG3. METHODS We studied 17 healthy volunteers and 28 patients with a wide range of renal function. For each plasma clearence curve, nine plasma samples were drawn at intervals from 10 to 240 min after injection of tracers. When comparing individual values for GFR and TER (MAG3) from the tracer dilution spaces (VD) with those derived from the analysis of the entire plasma disappearance curves of two radiopharmaceuticals, a good linear correlation appears (r = 0.96). RESULTS We found that the nadir-error (Sy,x) for predicted GFR occurs at 180 min (11.0 ml/min/1.73 m2), while the nadir-error for predicted TER (MAG3) is reached at 90 min (26.4 ml/min/1.73 m2). CONCLUSIONS In the computation of GFR and TER (MAG3) with a single-sample method, it appears that the mean residence time (t) for each tracer represents the optimum plasma sampling time. Our results suggest that the single injection of 51Cr-EDTA and 99mTc-MAG3 followed by blood sampling twice permits accurate simultaneous estimation of GFR and TER (MAG3) and, after correction of the latter kinetic parameter, effective renal plasma flow.