During a 6-year, hospital-based study at Newcastle upon Tyne five consecutive winter epidemics of respiratory syncytial (R.S.) virus infection occurred; the virus was identified in 1428 cases, showing that 1 in 50 live-births were admitted to hospital with R.S. virus infection. Epidemics were inversely related to temperature and to number of hours of sunshine. Parainfluenzaviruses, the second largest group of pathogens, were identified in 543 cases; most infections by this group were due to parainfluenzavirus type 3, which accounted for admission to hospital of 1 in 300 live-births. Epidemics of parainfluenza type 3 showed a summer peak (there was often a second peak in autumn) and a positive correlation with temperature and number of hours of sunshine. Epidemics of parainfluenza 1 and 2 occurred together, but only every 2 years. Influenza A epidemics occurred every winter; they coincided with, and had the same climatic correlations as, epidemics of R.S. virus infections. Influenza A was the second most frequently identified virus, and was associated with the admission to hospital of 1 in 100 to 1 in 500 live-births. Knowledge of the epidemiology of respiratory viral infections may help in the planning of preventative measures.