BACKGROUND: Although necessary for developing a rationale for vaccination, the burden of severe respiratory syncytial virus (RSV) disease in children in resource-poor settings remains poorly defined. METHODS: We conducted prospective surveillance of severe and very severe pneumonia in children aged <5 years admitted from 2002 through 2007 to Kilifi district hospital in coastal Kenya. Nasal specimens were screened for RSV antigen by immunofluorescence. Incidence rates were estimated for the well-defined population. RESULTS: Of 25,149 hospital admissions, 7359 patients (29%) had severe or very severe pneumonia, of whom 6026 (82%) were enrolled. RSV prevalence was 15% (20% among infants) and 27% during epidemics (32% among infants). The proportion of case patients aged 3 months was 65%, and the proportion aged 6 months was 43%. Average annual hospitalization rates were 293 hospitalizations per 100,000 children aged <5 years (95% confidence interval, 271-371 hospitalizations per 100,000 children aged <5 years) and 1107 hospitalizations per 100,000 infants (95% confidence interval, 1012-1211 hospitalizations per 100,000 infants). Hospital admission rates were double in the region close to the hospital. Few patients with RSV infection had life-threatening clinical features or concurrent serious illnesses, and the associated mortality was 2.2%. CONCLUSIONS: In this low-income setting, rates of hospital admission with RSV-associated pneumonia are substantial; they are comparable to estimates from the United States but considerably underestimate the burden in the full community. An effective vaccine for children aged >2 months (outside the age group of poor responders) could prevent a large portion of RSV disease. Severity data suggest that the justification for RSV vaccination will be based on the prevention of morbidity, not mortality.