Radioimmunoassayed serum concentration and urinary excretion data for digoxin from azotemic patients were characterized using a 2-compartment open model. Urinary excretion rates of digoxin as well as serum concentration data are needed to accurately characterize the disposition of the drug. Seven patients with renal failure showed highly variable steady-state volumes of distribution (V-ss-D equals 195 to 489 liters/1.73 m-minus2) and t1/2beta values (1.5 to 5.2 days). This variability is a major limiting factor in the use of dosage regimen nomograms that assume a constant V-ss-D and a rigorous relationship between t1/2beta and creatinine clearance (Cl-CR). Body clearance (Cl-B) is a parameter that is affected by both elimination and distribution of drugs. A linear relationship between Cl-B and renal clearance of digoxin or Cl-CR was found and was used to develop a model-independent approach to calculation of maintenance doses of digoxin. Several methods for calculating steady-state serum concentrations of digoxin (C-ss-p) were compared with actual measurements obtained in 16 chronically medicated patients. Optimum computation of C-ss-p is obtained by use of digoxin renal and body clearances. Variability in the digoxin:creatinine renal clearance ratio is the major limiting factor in prediction of digoxin dosage regimens.