To better understand acquired immunity to respiratory-syncytial-virus infections, we analyzed data from a 10-year study of respiratory illness in normal children who were followed longitudinally from early infancy. Immunity was measured in terms of failure to become infected or reduction in severity of clinical illness upon reinfection. Outbreaks of infections occurred seven times over the 10-year-period. During epidemics the attack rate for first infection was 98 per cent. The rate for second infections (75 per cent) was modestly reduced (P less than 0.001); that for third infections was 65 per cent. Age and history of infection both influenced illness. Immunity induced by a single infection had no demonstrable effect on illness associated with reinfection one year later; however, a considerable reduction in severity occurred with the third infection. These observations suggest that amelioration of illness--rather than prevention of infection--may be a realistic goal for immunoprophylaxis.