OBJECTIVES Available evidence on the effect of sickle cell trait (SCT) on birthweight is conflicting, not gestational age specific, and does not account for maternal and infant factors. The objectives of this study are to determine the contemporary mean birthweight, mean customised birthweight centile, and to analyse the risk of small-for-gestational-age (SGA) and large-for-gestational-age (LGA) babies in SCT pregnancies. DESIGN Large retrospective cohort study. SETTING London hospital. POPULATION Singleton pregnancies between 24 and 42 completed weeks delivered between 2000 and 2005 in parturient with body mass index between 18.0 and 35.0 kg/m(2). METHODS All qualifying pregnancies were identified on Terra Nova Healthware. Birthweight centiles of these cases were computed with Gardosi customised bulk centile calculator using collected data on maternal height, weight, ethnicity and parity, and the infant's gender, gestational age and birthweight. Birthweight and birthweight centiles of SCT and pregnancies with no haemoglobinopathy (control) were compared. Statistical analysis was performed using Stata version 9.2. MAIN OUTCOME MEASURES Birthweight and birthweight centiles. RESULTS Five hundred and five SCT and 16 320 controls were analysed. The mean birthweight of SCT pregnancies was 3223 g, 57 g lower than controls (P = 0.024). However, its mean birthweight centile was 49.0% similar to that of controls' 47.5% (P = 0.320). There is an apparent risk of LGA babies in SCT pregnancies, but logistic regression analysis suggests that the odds are related to being an older non-white parturient and a male infant rather than SCT status. CONCLUSIONS SCT is not a risk factor for SGA or LGA infants.