Since 1982, prospective information on smoking during early pregnancy is reported to the Swedish Medical Birth Registry for nearly all (93%) women who give birth. The present paper studies the validity of this information: effects on birthweight and perinatal mortality are very similar to those described previously in the literature. For each parity class, smoking decreases in inverse proportion to increasing maternal age; for each maternal age class, smoking increases with parity. A slight reduction in smoking rate is observed between 1983 and 1987, most pronounced for young women. There are marked geographic and social differences in the rate of smoking during pregnancy. This dataset can be used in the future to monitor the prevalence of smoking, and to study various factors associated with smoking and the impact of countermeasures taken against smoking during pregnancy. It can also be used to study possible associations between maternal smoking and rare events like congenital malformations and child cancer.