The safety and efficacy of rimantadine for long-term prophylaxis of influenza A (H3N2) infection were evaluated among elderly residents in 10 nursing homes. Within each nursing home, participating residents were randomly assigned to receive placebo or rimantadine at 100 or 200 mg/day. Residents were evaluated daily for symptoms and significant health events as possible side effects, as well as for influenza-like illness. The study medications were administered to 328 residents for up to 8 weeks, with no statistically significant differences in the frequencies of gastrointestinal or central nervous system symptoms between the groups. However, residents in the active medication groups were more likely to withdraw from the study and to experience various health events including death; some but not all of these differences were statistically significant. Efficacy evaluations were carried out on the 68 vaccinated residents in the two nursing homes with demonstrated influenza virus activity. Rimantadine appeared to provide an additional protective effect beyond vaccination in reducing the risk of clinical and laboratory-confirmed influenza-like illness; however, the efficacy estimates were never statistically significant. The efficacies of the 100- and 200-mg/day dosages were generally similar. When data for the 100- and 200-mg/day dosage groups were combined and compared with data for the group receiving placebo, the efficacy of rimantadine in reducing the risk of clinical influenza-like illness was estimated to be 58 percent (P = 0.079). The results suggest the relative safety and clinical efficacy of using rimantadine for influenza prophylaxis among vaccinated elderly individuals and support the recommendation for a dosage reduction to 100 mg/day in this population.