To assess the significance of the length of time between two pregnancies on the outcome of the second we used information collected by the British Perinatal Mortality Survey of 1958. From questionnaires on the 16,994 singleton births in the first week of March 1958 and the 7,117 singleton stillbirths and neonatal deaths in March, April, and May 1958 we abstracted information on the date and outcome of any preceding pregnancy. The interpregnancy interval was taken as the length of time between this preceding pregnancy and the last menstrual period before the index pregnancy. The most important factors influencing pregnancy spacing were outcome of the preceding delivery, social class, and maternal age. When these variables had been taken into account we found that the length of interpregnancy interval had little effect on stillbirth rates. High neonatal death rates, however, occurred when interpregnancy intervals were less than six months (P <0.005), though longer intervals had no significant effects.