We applied an open one compartment pharmacokinetic model for the determination of glomerular filtration rate (GFR) and effective renal plasma flow (ERPF) based on a rapid intravenous loading dose followed by a constant infusion of 125I-iothalamate and 131I-orthoiodohippurate in order to ensure constant plasma levels of the two clearance markers. The loading dose was based on the assumption that the volume of distribution of the two markers equals the extracellular volume (25% of the body weight). The infusion rate as calculated after the clearance of thalamate was estimated from body weight, age, sex and serum creatinine using Cockcrofts formula. The clearance of hippurate was assumed to be four times that of thalamate. We studied the reliability of this model in 212 patients with insulin dependent diabetes mellitus (IDDM; n = 74), nephrotic syndrome (NS; n = 18) and heart (HTX; n = 69) or kidney (KTX; n = 51) transplants. A steady state concentration was obtained in all patient groups, even when GFR was markedly depressed. In patients with diabetes, we observed more variance between plasma and urinary clearances of thalamate, which could be due to inaccuracies in urine sampling. In these patients, GFR should be measured using a method that is not dependent on urine collection. Also, the estimation of GFR by means of Cockcrofts equation seems to underestimate GFR in diabetic subjects.