Because noninstitutionalized senior citizens comprise over 95% of the population 65 years of age and older, their health needs are a major concern. Data regarding infections in this population including the epidemiology, morbidity, and mortality are lacking. The authors recruited a study population of 417 free-living persons, all 65 years of age or older, from two neighborhoods in Pittsburgh, Pennsylvania. After the collection of self-reported baseline information from these persons, they were monitored for all clinical infections for 2 years, beginning July 1986 and through June 1988, using clinic visits, hospitalizations, or phone calls when needed. The baseline information showed the study population of 417 persons to be comparable with a neighborhood comparison group and with established populations for epidemiologic studies of the elderly in three other states. The 24 months of infection surveillance yielded 494 diagnosed infections in 224 or 54% of the subjects. Respiratory infections were most frequent with 259 or 52% of the total, followed by genitourinary infections with 24%, skin infections with 18%, gastrointestinal infections with 4%, and other types of infection with 2%. By comparing 22 self-reported baseline conditions with the occurrence of infection, 10 historic factors were univariately significant for infection. Of these 10 factors, only history of a lung problem (relative risk = 1.7, 95% confidence interval (CI) 1.1-2.9) and history of difficulty controlling urination (relative risk = 2.7, 95% CI 1.3-4.9) were statistically significant in multivariate analysis. To our knowledge, this study represents the first prospective data on infections in the noninstitutionalized elderly. The data demonstrate the wide variety of infections that occurred in this population and suggest that persons with a history of any one of several medical problems were possibly at greater risk for infection.