An outbreak of influenza occurred at the end of the 2001-2 winter season in a highly-vaccinated, semi-enclosed, religious community. On the basis of retrospective self reporting, 43% met the case definition (151/350) and 81% (25/31) of throat swabs from cases were polymerase chain reaction (PCR) -positive for influenza A (H3N2). The risk of developing influenza in people aged 65 or more was lower than that of children aged 2 years and under (odds ratio 0.1, 95% confidence interval 0.02 to 0.38). The risk of developing symptoms of influenza was not significantly different between people who had been vaccinated in the United Kingdom and those who had not been vaccinated (OR 1.14, CI 0.41 to 3.14). Chronic disease was an independent risk factor for developing symptoms of influenza (OR 1.9, CI 1 to 3.63). Timing of the outbreak, the age structure and mode of communal living may have influenced the efficacy of the influenza vaccine, which was well matched to circulating strains at the time of the outbreak. It is important to consider the diagnosis of influenza even in a highly-vaccinated community.