In order to assess the practical reliability of glomerular filtration rate (GFR) determination with 99mTc-DTPA and plasma sampling, the authors compared the results obtained with 51Cr-EDTA and 99mTc-DTPA in 50 patients using five easily applied methods (two double-plasma-sample methods and three single-plasma-sample methods), and two kits with different compositions. It was observed that: 1) there is no difference between the results obtained with the two different kits. 2) Compared with 51Cr-EDTA the 99mTc-DTPA overestimates the result by about 2 mL/min: precision is slightly lower with 99mTc-DTPA than with 51Cr-EDTA but is sufficient for practical use. 3) The method recommended by the authors on the basis of this experience is the Russell's method with two samples. 4) The simplified methods with one sample give comparable results to the Russell's method for GFR levels between 50 and 115 mL/min, while the results are unsatisfactory below 50 mL/min. 5) Among the single-sample methods, the authors suggest that of Christensen and Groth. 6) A preliminary estimate of GFR (from the serum creatinine level, for instance) is useful for the choice between double-plasma-sample methods and simplified methods. In conclusion, the authors consider that the estimation of GFR with 99mTc-DTPA can be performed efficiently in clinical practice even when operating in absolutely routine conditions.