Smoking during pregnancy is causally associated with reduced birth weight and is strongly related to preterm birth. Smoking cessation in early pregnancy seems to reduce these risks, although the research evidence is limited. In a sample of Romanian women, differences in birth outcomes were assessed between non-smokers and women who continued to smoke during pregnancy and non-smokers and women who stopped smok- ing when they found out about the pregnancy. Pregnant women were recruited in two urban clinics (N= 474). A baseline questionnaire collected information on their smoking status, depressive symptoms, stress, demographics, and other characteristics at recruitment. The women reported the newborn weight and birth term by phone in the first weeks following birth. Descriptive statistics and multivariate regressions were used to ana- lyze the relationship between smoking status during pregnancy and birth outcomes. Over 61% (N = 290) women were non-smokers, 15% (N= 72) smoked during pregnancy, and 24% (N= 112) quit smoking when they found out about the pregnancy. Compared to non-smokers, continuous smokers delivered babies 165 grams lighter (95% CI -313, -17). Women who stopped smoking when they ascertained the pregnancy had higher odds of delivering a newborn who was small for gestational age compared to non-smokers (OR= 2.16, 95% CI 1.05, 4.43). Elevated maternal stress was associated with reduced birth weight (-113 grams, 95% CI -213, -11), and higher odds of a preterm birth (OR=2.8, 95% CI 1.17, 6.76). In a predominantly urban sample of Romanian women, continuous maternal smoking during pregnancy was a risk factor for restricted foetal growth. Smoking cessation when the pregnancy was ascertained did not seem to reduce this risk. Smoking prevention efforts should therefore begin before pregnancy and should integrate psychological components, addressing maternal stress in particular.