This study investigates whether a child's risk of rotavirus diarrhoea is associated with season of birth in England and Wales, countries where rotavirus infections are highly seasonal. Poisson regression models were fitted to weekly counts of laboratory-confirmed rotavirus infections from children aged <5 years born between 1998 and 2007. In the first year of life, the risk of a laboratory-confirmed rotavirus infection was significantly higher for children born in summer compared with winter [relative risk (RR) 2·13, 95% confidence interval (CI) 2·07–2·19]. In the second to fifth years of life, the pattern reversed (second year of life: RR 0·73, 95% CI 0·71–0·75). The cumulative risk up to age 5 years remained significantly higher for children born in summer compared with winter due to the increased risk for summer births during their first year of life. Maternal immunity and age-specific levels of exposure to rotavirus could explain our findings.