Background To examine if smoking during pregnancy is associated with poorer motor competence among offspring, indicating impaired neurological function. The measures may be less susceptible to socioeconomic confounding than cognition tests.

Methods Data were from 13 207 members of the National Child Development Study, born in Great Britain in 1958. Maternal smoking during pregnancy was recorded prospectively. Tests of physical control and coordination administered by a school doctor at age 11 years were: time to pick up 20 matches (PUM), number of squares marked (NSM) and copying designs (CD). PUM and NSM were tested for left and right hand. Test scores were dependent variables in linear regression analysis, with adjustment for maternal smoking during pregnancy, sex, birth weight standardised for gestational age, breast-feeding, social class, parental education, mother's age, laterality and pubertal development.

Results After adjustment, heavy smoking during pregnancy was statistically significantly associated with PUM (non-dominant hand) and CD, but not NSM; particularly among boys. The regression coefficients (and 95% CI) for PUM (non-dominant hand) are 1.474 (0.47 to 2.48, p=0.004) and 1.203 (0.15 to 2.26, p=0.026) for boys and girls, respectively: higher scores indicate poorer performance. The coefficients for CD are −0.185 (−0.32 to −0.05, p=0.006) for boys and 0.020 (−0.11 to 0.15, p=0.753) for girls: lower scores indicate poorer performance.

Conclusions Smoking during pregnancy is associated with subtly reduced motor competence, particularly on the non-dominant side. Statistically significant effect modification by sex was observed for only one test, providing equivocal evidence of a sex difference.