BACKGROUND:: Live attenuated influenza vaccine (LAIV; FluMist) is a trivalent vaccine containing cold-adapted influenza vaccine viruses that infect and replicate in cells lining the nasopharynx to induce immunity. Recovery of viruses (shedding) is measured by culture of nasal specimens. Shedding of vaccine viruses is not equated with transmission because transmission requires more virus than is detected in many nasal swabs. Previous studies with LAIV did not detect transmission to close contacts. The primary objective of this study was to estimate the probability of transmission to placebo contacts in a day care setting. METHODS:: One hundred ninety-seven healthy children aged 9 to 36 months attending day care were randomized to receive vaccine or placebo. Postvaccination viral shedding, safety, genotype and phenotype of shed viruses and probability of transmission were assessed. RESULTS:: Eighty percent of 98 vaccine recipients shed at least one vaccine strain. No clinically significant differences in solicited adverse events attributable to vaccine occurred; safety profiles were similar in both groups. Vaccine virus isolates retained their phenotypic characteristics (cold adaptation and temperature sensitivity) and did not revert at nucleotides known to confer an attenuating phenotype. There was one confirmed transmission of a vaccine strain to a single placebo recipient. According to the Reed-Frost model, the calculated probability of transmission to a child after contact with a single vaccinated child was 0.58% (95% confidence interval, 0-1.7%). There was no increased reactogenicity or other safety concerns in the recipient child. CONCLUSIONS:: Young children in a day care setting had a high rate of shedding and a low rate of transmission. No clinically significant illness occurred among children who received vaccine or placebo or in the child to whom the vaccine virus was transmitted.