An outbreak of influenza A in nursing home residents is reported and other studies of influenza vaccine effectiveness in elderly populations are reviewed. The outbreak occurred in a Los Angeles nursing home between February and March 1983. Of the 87 residents, 46 (53%) were affected with influenza-like illness. Attack rates were similar between immunized and unimmunized residents (52% versus 58%), and yielded a vaccine effectiveness estimate of 10%. No additional protection could be demonstrated in residents who received vaccine for two consecutive years. Seven persons died (mortality rate of 8.1%); the mortality rate was greater in the unimmunized (15.8%) than in the immunized (6.2%). Because this study and other field studies of influenza vaccine are limited in precision and power, a statistical summary of the various studies was constructed. Summarizing the studies of institutionalized elderly (in hospitals and nursing and retirement homes) yielded an estimate of 74% for the average vaccine effectiveness in mortality reduction, and an estimate of 33% for the average effectiveness in preventing clinical illness. For the non-institutionalized elderly, the corresponding estimates were 47% for mortality, and 5% for clinical illness. Despite the obvious limitations of such summaries, it seems reasonable to conclude that influenza vaccines have on the average been of clear benefit in the institutionalized elderly, while the benefits in the non-institutionalized elderly have been less dramatic and may warrant further investigation.