A population pharmacokinetics study using the NONMEM program was undertaken to determine the effects of different covariates on the pharmacokinetic parameters of etoposide. A total of 1,044 plasma etoposide concentrations were determined by high-performance liquid chromatography (HPLC) in 100 patients (pts; 75 men and 25 women aged 25-85 years) treated for various tumor types with i.v. (57 pts) or oral (43 pts) etoposide. For 67 pts, etoposide plasma protein binding was determined by equilibrium dialysis; the unbound fraction ranged from 4% to 24%. A linear two-compartment model with first-order absorption (for oral dosing) accurately described the concentration versus time data. The central and peripheral volumes of distribution were significantly correlated with the body surface area [Vc (L) = 5.5 x BSA (m2) and Vp = 4.1 x BSA], but even after BSA had been taken into account, the interindividual variability of the two volumes remained high (34% and 57%, respectively). The clearance (CL) was not correlated with the following covariates: age, BSA, sex, height, and levels of serum bilirubin and liver enzymes. The final regression model for CL was CL (ml/min) = 49.8 x (1 - 0.009 x PRO) x WT/Scr + 33.8 x (1 - 0.29 x META) x (1 - 0.012 x ALB), where ALB, PRO, WT, and Scr, respectively, were albuminemia, proteinemia (g/l), weight (kg), and serum creatinine (microM) and META = 1 if the patient had liver metastases (otherwise, META = 0). The interindividual variability in CL (mean value 30 ml/min) decreased only from 32% to 26% when these covariates were taken into account. The mean oral bioavailability was 66%, showing an interindividual variability of 37%. The plasma clearance of the unbound fraction was strongly and negatively correlated with Scr but was not dependent on either PRO or ALB. These data show that modifications in PRO levels do not directly affect plasma exposure to unbound etoposide. This analysis makes possible the rational consideration of modifications of covariates such as Scr in etoposide dosing. This population data base will constitute the prerequisite for adaptative control with feedback dosing for continuous oral administration of etoposide.