Zanamivir is a potent and specific inhibitor of influenza A and B virus neuraminidase, that is now approved for the treatment, and is currently under development for the prophylaxis of influenza. To assess the safety of this drug in asthmatics, 13 subjects with mild/moderate asthma [forced expiratory volume in 1 sec (FEV1)> or =70% predicted, reversibility of FEV1 to salbutamol > or =15%, concentration of methacholine causing a drop of 20% in the FEV1 (PC20FEV1)< or =8 mg ml(-1)], were recruited to a double-blind, randomized, placebo controlled, two way cross-over study. Subjects received 10 mg zanamivir as a dry powder (2 x 5 mg blisters via a Diskhaler Sovnn Plastics Ltd., Berkshire, U.K.), or a matching placebo, twice daily on day 1 and then four times daily from day 2 to day 14, in two separate periods separated by a washout period of 7 days. PC20FEV1 to methacholine was determined pre-study, on day 1 after the evening dose and on day 14 after the last dose of the study drug. FEV1 was measured pre-study and at regular intervals on days 1 and 14. Laboratory safety tests were performed on days 1, 7 and 15. Morning and evening peak expiratory flow rate (PEFR) and any adverse events were recorded in a diary card. Eleven subjects completed the study. One was withdrawn due to non-compliance, and one due to an adverse event that occurred during the placebo period. On day 1 the geometric mean PC20 for zanamivir was 36% lower than for placebo [ratio to placebo 0.64, (90% CI 0.44, 0.93)] and on day 14 this was 33% lower with zanamivir [ratio to placebo 0.67 (90% CI 0.38, 1.15)]. Both these confidence intervals were within the pre-defined interval of 'no clinically significant effect' of 0.25-4 (i.e. a change of two doubling doses of methacholine PC20FEV1 which was considered clinically significant). The time weighted mean FEV1 was 0.15 l (5.4%) lower for zanamivir on day 1 compared to placebo (90% CI 0.03, 0.28; P=0.050) and 0.01 l higher compared to placebo on day 14 (90%CI -0.12, 0.10; P=0.912). The day 1 changes were not associated with any significant symptoms or requirement for rescue bronchodilator therapy. Furthermore there was no apparent treatment difference over the 14 day dosing period in FEV1 data (90% CI: -0.11, 0.05, P=057). The mean morning PEFR was 4 l min(-1) less for zanamivir than for placebo (90% CI: -11, 3) and mean evening PEFR was 9 l min(-1) less (90% CI: -24, 5). The study treatments were well tolerated by the subjects with no clinically significant adverse events attributable to zanamivir treatment. Zanamivir inhaled as a dry powder does not significantly affect the pulmonary function and airway responsiveness of subjects with mild/moderate asthma and therefore its use in such patients subjects is not precluded.