The analysis of a survey of 3759 births shows that the difference in placental weight between smokers and nonsmokers is, if not nil, at least minimal, and that is definitely less than might be expected taking birth weight and other factors into consideration in a multivariate analysis (Tab. II). The distortion between the fetal and placental weights, which are very closely correlated, assigns to smoking a peculiar place among the factors influencing these two weights, a result which may lead to various interpretations.