Using a standardised procedure, 20 women who were 37 weeks pregnant logged fetal movements for 1 h in the morning and evening for 3 days. Close agreement between such logs and ultrasound recordings of individual differences in fetal movements have been documented. After birth, the babies' sleeping, waking, feeding and crying behaviour were recorded prospectively in validated behaviour diaries at 1, 6 and 12 weeks of age. The fetal movements showed an evening peak, due to a diurnal variation in strong movements. These strong movements did not predict the babies' crying. Weak fetal movements provided a reliable index of fetal differences and predicted the amounts the babies cried at 1, 6 and 12 weeks of age. The findings suggest that temperament-like dispositions present before babies are born predict how much they are likely to cry.