Background:Spontaneous preterm delivery (sPTD) is one of the most common global problems causing perinatal morbidity and mortality. However, there is no effective universal screening method for predicting sPTD. Therefore, we want to evaluate the extra benefit of quadruple maternal serum biomarkers for Down syndrome (DS) screening to predict sPTD. Moreover, to propose an effective model to predict sPTD.Results: From total of 2,822 pregnancies who underwent DS screening, 2604 (92.3%) were included in the study. There were 183 (7.02%) pregnant women had randomly measure cervical length. sPTD occurred in 221 (8.5%) of participants, 32 (14.5%) had early sPTD (<34 weeks gestation) and 189 (85.5%) had late sPTD (delivery between 34-36+6 weeks). The median MoM of AFP was significantly higher in women with early sPTD (1.2 VS 1.0, p =0.04), but uE3 had tendency to be lower than the women with term delivery. The most effective model to predict sPTD was the combination of serum AFP level > 1.2 MoM, uE3 level < 0.9 MoM, cervical length <35 mm. and had history of previous preterm delivery which has an AUC of 0.86.Conclusions:An abnormal increase of AFP or decrease of uE3 in quadruple Down syndrome screening showed a predictive ability to predict sPTD with no extra cost. The most effective model to predict sPTD is adding serum quadruple biomarkers with maternal characteristics and cervical length measurement gives a high predictive ability for sPTD.