Objective: To construct a prediction model for preeclampsia (PE) risk in twin-pregnant women, so as to provide the basis for early screening and prevention of PE. Methods: A total of 467 twin-pregnant women who underwent prenatal examination and delivered at Huzhou Maternal and Child Health Hospital were selected. Sixty cases with preeclampsia (PE) were included in the case group, and 60 women without PE were included in the control group. General information, blood biochemical indicators and uterine artery resistance index (UtA-RI) were collected. A logistic regression model was used to screen predictive factors and establish a nomogram. The Bootstrap method was performed for the internal validation; the receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis were employed to evaluate the discrimination, calibration and clinical utility of the nomogram, respectively. Results: In the case group, there were 47 individuals (78.33%) aged younger than 35 years, 21 individuals (35.00%) with pre-pregnancy body mass index (BMI) of 25 kg/m2 and above, and 33 individuals (55.00%) with in vitro fertilization. In the control group, there were 57 individuals (95.00%) aged younger than 35 years, 8 individuals (13.33%) with pre-pregnancy BMI of 25 kg/m2 and above, and 39 individuals (65.00%) with natural pregnancy. Multivariable logistic regression analysis identified age, pre-pregnancy BMI, method of conception, placental growth factor (PLGF) and UtA-RI as risk prediction factors for PE in twin-pregnant women. The established nomogram had an area under the ROC curve of 0.827 (95%CI: 0.755-0.899), a sensitivity of 0.767, a specificity of 0.733, a good discrimination and calibration, and a relatively high clinical net benefit. Conclusion The nomogram established by age, pre-pregnancy BMI, method of conception, PLGF and UtA-RI has a good predictive value for the risk of PE in twin-pregnant women.