A double-blind, randomized study of inhaled zanamivir for the prevention of influenza in families was conducted. Once a person with a suspected case of influenza was identified (index patient), treatment of all other household members (contacts) > or =5 years old was initiated. Contacts received either 10 mg zanamivir or placebo inhaled once daily for 10 days. Index patients received relief medication only. In total, 487 households (242 placebo and 245 zanamivir) were enrolled, with 1291 contacts randomly assigned to receive prophylaxis. Four percent of zanamivir versus 19% of placebo households (P<.001) had at least 1 contact who developed symptomatic, laboratory-confirmed influenza, representing 81% protective efficacy (95% confidence interval, 64%-90%). Protective efficacy was similarly high for individuals (82%) and against both influenza types A and B (78% and 85%, respectively, for households). Zanamivir was well tolerated and was effective in preventing influenza types A and B within households where the index patient was not treated.