Epidemiological data, especially population-based data, on respiratory syncytial virus (RSV)-related hospitalizations in Germany have been lacking to date. Since Palivizumab (Synagis, Abbott, USA) is already available and new vaccines for active immunization are under development, these data are urgently needed. From July 1996 to June 1999, nasopharyngeal aspirates of children hospitalized in Kiel with an acute respiratory tract infection were tested by multiplex reverse transcriptase polymerase chain reaction. Of 1,241 patients, 150 (12.1%) were RSV positive. RSV was the predominant pathogen detected through the end of the second year. In 37 (25%) children an underlying condition was present. For the city of Kiel and close surroundings, the cumulative incidence of RSV-positive hospitalizations in infants was 1,214/10(5) (725/10(5) in children less than 2 years). For those children less than 2 years old born with a gestational age of less than 32 weeks or 32-37 weeks, the cumulative incidence was 2,025/10(5) and 1,202/10(5), respectively (dose-effect response). For the group less than 32 weeks of age, bronchopulmonary dysplasia (BPD) as an underlying condition carried a relative risk of 17.8. The RSV season began between the end of September and January and ended between March and July. The population-based incidence of RSV-positive hospitalizations in Kiel is close to that reported from the UK and Scandinavia. Throughout Germany, approximately 10.000 RSV-related hospitalizations in infants can be expected annually. Prematurity is an effect modifier and BPD a strong risk factor for RSV-positive hospitalization in population-based studies. There is considerable variation in the start and end of the yearly epidemic.