The pharmacokinetics of zanamivir were evaluated in subjects from three phase I single-center and two phase II multicenter, randomized, double-blind, multidose, placebo-controlled trials. A total of 96 phase I subjects received zanamivir (3.6 to 16 mg) intranasally two or six times daily for 4 to 5 days beginning 4 hours before or 1 to 2 days after inoculation with influenza virus. A total of 75 phase II subjects with influenza or a history of exposure to naturally occurring influenza virus were administered zanamivir as an intranasal spray (3.4 mg/nostril), inhaled powder (10 mg), or combination of intranasal and inhaled formulations twice daily for 5 days. Population parameters (including demographic factors, zanamivir formulation, infection-related variables, and concurrent medication use) were estimated by a nonlinear mixed-effect modeling software program (NONMEM) using a one-compartment model with first-order absorption and conditional estimation algorithm. Formulation and route of administration were the most significant factors affecting the pharmacokinetics of zanamivir. Relative bioavailability of the inhaled powder to the intranasal drops and spray was 2.3 and 1.6, respectively. No significant differences in pharmacokinetic parameters were observed when demographic variables, indices of infection, or concurrent medication use were considered in either phase I or phase II population analyses.