The objective of this study was to accurately determine the within- and between-patient variability in etoposide pharmacokinetics for i.v. and p.o. administered drug. Inter-and intrapatient variability in systemic etoposide exposure was measured following i.v. and p.o. drug administration using stable isotope dilution methodology. Seven patients received 50 mg of etoposide by both p.o. and i.v. routes of administration on three separate occasions 1 month apart. Etoposide plasma concentrations following p.o. and i.v. drug administration were quantitated by liquid chromatography-mass spectrometry for each route of administration. The area under the plasma etoposide concentration versus time curve, plasma etoposide clearance, and etoposide plasma half-life were calculated for each dose of drug. Kinetic measurements following i.v. and p.o. drug administration were compared. The within-patient variation in the areas under the plasma etoposide concentration versus time curves following i.v. drug administration was minimal [coefficient of variation (CV) = 9.3%]. Within-patient variability was increased 2.4-fold with oral drug administration (intrapatient CV = 22.2%). Between-patient variability was roughly three times as great as within-patient variability (interpatient i.v. CV = 28.4%; interpatient p.o. CV = 58.3%). Mean etoposide bioavailability at a dose of 50 mg was 64.6%, again with greater interpatient than intrapatient variability (34.8 versus 22.6%). Greater variation in drug toxicity is expected with p.o. compared with i.v. etoposide use. Administration of repeated doses of etoposide to the same patient should produce less variation in toxicity than between-patient dosing.