The aim of this paper was to determine the reproducibility (precision) of technetium-99m mercaptoacetyltriglycine (MAG3) clearance and to compare it with that of chromium-51 ethylenediamine tetra-acetic acid (EDTA). Twelve young volunteers (aged between 21 and 34 years), without any history of medical problems, were enrolled in this study. The test was performed twice, at an interval of 8 days and under similar physiological conditions. After the intravenous injection of both tracers, 15 blood samples were taken between 3 and 240 min. A biexponential fit was adapted to the plasma disappearance curves (5- to 120-min samples for 99mTc-MAG3 and 10- to 240-min samples for 51Cr-EDTA); the clearances were calculated according to Sapirstein and corrected for body surface area. The mean clearance values were 110 (range 85-130) and 226 (range 109-319)ml/min/1.73 m2, respectively, for 51Cr-EDTA and 99mTc-MAG3. For 51Cr-EDTA clearance, the mean difference between the first and the second measurement was -2.1% of the mean of the two successive values (SD: 8.4%). In ten cases, the difference was less than 12%; in two cases, the differences were 15% and 18%, respectively. For 99mTc-MAG3 clearance, the mean difference between the first and the second measurement was -20% of the mean of the two successive values (SD: 25%). In six cases, the difference was less than 12%; in four cases, between 15% and 40%; and in two cases, more than 60%. Methodological factors (impurities contained in the commercial kit, variable protein binding) as well as physiological factors (pH of urine, sodium load, stress) may explain the lack of precision of the 99mTc-MAG3 clearance. It is concluded that changes in 99mTc-MAG3 plasma clearance should be interpreted with care in daily routine.