OBJECTIVE smoking is an important risk factor for placental abruption with strong dose-dependency. Pregnant smokers often underreport tobacco use which can be objectively assessed by measuring serum cotinine levels. We examined the accuracy between self-reported smoking habits and early pregnancy serum cotinine levels in women with or without placental abruption. METHODS retrospective case-control study. METHODS university Hospital. METHODS a total of 175 women with placental abruption and 370 control women. METHODS serum samples collected during the first trimester were analyzed for serum cotinine levels. Cotinine concentration over 15 ng/ml was considered as the cutoff indicating active smoking. Smoking habits of the women and their partners were recorded at the same visit. METHODS placental abruption. RESULTS of the cases of women with placental abruption, 27.4% reported smoking compared with 14.3% of the controls (p < 0.001). Based on serum cotinine levels, 30.3% of the case women and 17.6% of the control women were considered smokers (p = 0.003). Serum cotinine levels among smokers were higher in the abruption group than in the control group (median 229.5 ng/ml (interquartile range 169.8-418.1) vs. 153.5 ng/ml (56.6-241.4), p = 0.002). Self-reported number of cigarettes smoked daily correlated well with the cotinine levels (r = 0.68, p < 0.001). Of the women reporting as nonsmokers, approximately 7% were considered smokers based on cotinine testing. CONCLUSIONS pregnant women with subsequent placental abruption are heavier smokers than pregnant control women. Self-reported smoking habits correlate well with serum cotinine levels in Finland. Therefore, self-reported smoking can be considered as a risk marker for placental abruption.