A radioimmunoassay was used to measure vincristine sulfate concentrations in the serum of four children with malignancies (ages 5-16 years) following intravenous (IV) bolus injections. The pharmacokinetic data were analyzed by a non-linear least-square regression program NONLIN. A three-compartment open model fitted the raw data better than a two-compartment model in three patients. In the other patient the raw data fitted a two-compartment open model. The half-lives of the triphasic decay curves alpha, beta, and gamma were 2.6, 41, and 1,531 min (25.5 h), respectively. The mean apparent volume of the central compartment was 3.25 l, and the the volume of distribution per 1.73 m2 body surface area at steady state was 215.9 l. In a three-compartment open model, the first-order distribution and elimination rate constants (min-1) of vincristine were as follows: k12, 0.088; k13, 0.121; k21, 0.028; k31, 0.0026; k10, 0.045. The plasma clearance was 146.2 ml/min per 1.73 m2, while the AUC0 infinity was 27,816 nM . min. Urinary excretion in one patient demonstrated a drug concentration of greater than 1.0 X 10(-7) M in the urine up to 78 h after the injection. Up to 37% of the administered drug was excreted in the urine as vincristine and/or its metabolites by 90 h. The low elimination rate constant from poorly perfused tissues to blood plasma (k31), a large apparent volume of distribution, and a long biological half-life (25.5 h) indicate avid tissue binding from which a slow release of the drug from the body tissues occurs.