Eleven patients with advanced cancer received navelbine (15 mg/m2) as a single i.v. bolus injection. At least 1 week later, the patients were given a 2-fold increased dose of navelbine (30 mg/m2) and, for seven of them, the 30-mg/m2 dose was repeated after a delay longer than a week. After each administration, plasma and urine were collected for 72 h and monitored for navelbine concentration by radioimmunoassay. The comparison of dose-normalized plasma level profiles showed significant time dependence (P less than 0.05) in four of the seven assessable patients. Some patients also exhibited significant (P less than 0.05) nonlinear (dose dependent) kinetic profiles. Only 3 of the 10 appreciable patients were characterized by both time independent and linear profiles. However, the plasma concentration decay curves presented a triphasic shape similar to that obtained with other antitumor Vinca alkaloids and the data were consistent with a three-compartment pharmacokinetic model. The dose and/or time dependence evidenced for most of the patients did not result in marked changes in pharmacokinetic parameters among courses. The pharmacokinetics of navelbine were characterized by a high plasma clearance (0.27 to 1.49 liter.h-1.kg-1), a large distribution volume (8.2 to 48.2 liter.kg-1), and a long terminal half-life (22.1 to 67.8 h). Urine excretion was low (less than 7.9%). Thus, navelbine pharmacokinetics resembles that of other antitumor Vinca alkaloids.