Maternal smoking has long been identified as a risk factor for sudden infant death (SID). However, only few studies analysed the biological plausibility of the relationship between maternal smoking and SID. In Lower Saxony (North Germany), detailed information concerning the perinatal period is routinely obtained for almost all infants born in this region. The perinatal data sets from 190 SID cases who had died between 1986 and 1990 and in whom a full autopsy had been performed were identified and compared to data sets from 5920 random controls, frequency matched to cases on year of birth. After adjusting for potential confounders (socio-economic status, birth weight, maternal age and nationality), smoking during pregnancy was still associated with a significantly increased risk of SID (odds ratio (OR) 2.7, 95% confidence interval (CI) 1.7-4.5). There was a clear dose-effect relationship between the number of cigarettes smoked and the risk of SID: adjusted ORs were 2.6 (1.5-4.4) for 1-10 cigarettes/day, 2.8 (1.8-6.0) for 11-20 cigarettes/day, and 6.9 (1.9-25.5) for > 20 cigarettes/day. There also appeared to be an interaction between smoking during pregnancy and maternal anaemia: the risk of SID almost doubled if mothers not only smoked, but were also anaemic (haemoglobin < 100 g/l). These results support the concept that smoking during pregnancy has direct biological effects on the fetus which are associated with an increased risk of SID later in life. The exact mechanism(s) whereby smoking increases the risk of SID, however, remains to be determined.(ABSTRACT TRUNCATED AT 250 WORDS)