Objective To investigate possible multigenerational influences on birthweight.

Design Data from the longitudinal study of one week's births in 1958 up to the age of 23 years, the British National Child Development Study, were utilized. These pro-vide socio-biological information on the parents of the cohort, on the cohort members from birth onwards, and on the pregnancies and the birthweight of any babies born to the cohort members.

Main outcome measure The main outcome was the birthweight of babies born to the cohort members, for whom complete intergenerational data were available for 1638 firstborn. Multiple regression modelling was used to investigate any associ-ations between their birthweight and characteristics of their parents and grandparents.

Results Significant positive associations were found between babies' birthweight and parental birthweight but not gestational age. For the babies born to female cohort members additional findings included associations between their birthweight and the height of the maternal grandmother and the social class of the maternal grandfather, even after adjustment for such strong predictors of birthweight as maternal weight, smoking habit in pregnancy and baby's sex and birth order.

Conclusion These results thus offer support for a multigenerational influence on birthweight passed through the maternal line.