Mean blood pressure (MBP), bilateral notching (BN) in the uterine artery and increased circulating levels of soluble fms-like tyrosine kinase-1/placental growth factor (sFlt-1/PlGF) ratio are predictors of preeclampsia (PE). Recently, we disclosed that reducing the plasma level of hydroxysteroid (17-b) dehydrogenase 1 (HSD17B1), which is a steroidogenetic enzyme catalyzing the conversion of estrone to 17b-estradiol, is a potential prognostic factor for PE. Our aim was to evaluate whether HSD17B1 is an independent risk factor for predicting PE after adjusting for the effects of MBP, BN and the plasma level of the sFlt-1/PlGF ratio in the second trimester. One hundred and twenty-eight consecutive normal pregnant women without gestational hypertension (GH) or PE and 30 women with PE were selected from 1724 pregnant women. Multivariate logistic regression with a forward stepwise procedure was used to construct a prediction model. A past history of GH/PE, a family history of hypertension, pre-pregnancy body mass index, MBP, BN, plasma levels of sFlt-1/PlGF ratio and plasma levels of HSD17B1 were significantly associated with the occurrence of PE; however, only MBP (OR (95% confidence interval), 1.08 (1.03-1.14)), BN (7.5 (1.9-30)), sFlt-1/PlGF (21 (2.7-163)) and HSD17B1 (0.43 (0.22-0.85)) were independent risk factors for PE. The area under the receiver-operating-characteristic curve for the combination model was 0.89, yielding a sensitivity of 0.84, a specificity of 0.88 and a positive likelihood ratio of 7.2 (4.0-13). In conclusion, HSD17B1 is an independent risk factor for PE, and the combination of several risk factors including HSD17B1 in the second trimester may improve the prediction of PE.