The precision of two different clearance methods as used for routine assessment of the glomerular filtration rate (GFR) was investigated in thirty-one children aged 0.6-14 years: total [51Cr]EDTA plasma clearance (E) determined by a simplified single injection method; and 24 h endogenous creatinine clearance (C). Determination of C twice only succeeded in twenty children because of problems in collecting 24 h urine accurately. The precision (determined from the total day-to-day variation) for single determinations in patients with E greater than or equal to 80 ml/min/1.73 m2 was 5.5% for E and 13.8% for C. The corresponding figures for E less than 80 ml/min/1.73 m2 was 7.8% for E and 20.8% for C. Data in the literature on the inaccuracies of C and E versus GFR suggest that the degree of inaccuracy in predicting GFR from C is much higher than that from E, a feature which together with the present findings on precision indicates that E is much more reliable than C for routine determination of GFR in children.