The purpose of the paper was to compare the effects of active and passive tobacco smoking during pregnancy on the birth weight of newborns and to validate self-reported smoking habit against the plasma cotinine level. Investigations on birth weight have been carried out in the sample of 1165 schoolchildren. Data on tobacco smoke and the birth weight of children were collected by standardized interviews with mothers. It was disclosed that the effect of passive smoking in never smokers was weaker (weight reduction by 73.1 g) than that of active smoking. It was estimated that active tobacco smoking of 10 or more cigarettes per day is to reduce the birth weight by about 450 g. Both effects of active and passive smoking in pregnancy were statistically significant. When the self-reported smoking status was compared with biochemical marker of exposure (plasma cotinine > 10 ng/ml) in 158 women at delivery, it became apparent a low sensitivity (43%) of the self-reported data on smoking, however, combined with high specificity (95%). It was shown that the exposure bias lead to a significant underestimation of association between low birth weight and tobacco smoking.