Objective To analyze risk factors associated with fetal growth restriction (FGR) in pregnant women with gestational diabetes mellitus (GDM) and polycystic ovary syndrome (PCOS), and to establish an FGR risk prediction model based on metabolic, endocrine, and placental blood flow characteristics. Methods A retrospective study was conducted including 100 pregnant women with GDM and PCOS who delivered at Anxi County Maternal and Child Health Hospital between January 2022 and January 2025. FGR was de?ned as estimated fetal weight (EFW) below the 10th percentile (P10) for gestational age. Participants were divided into an FGR group (n=28) and a non-FGR group (n=72). The differences in general characteristics, metabolic indicators, endocrine levels, pregnancy complications and placental blood ?ow parameters were compared between the two groups. Univariate and multivariate Logistic regression analysis was employed to identify independent risk factors for FGR. A predictive model was constructed based on the final model, with model performance evaluated using receiver operating characteristic (ROC) curves and calibration curves. Results The incidence of FGR was 28.0%. The FGR group exhibited lower pre-pregnancy body mass index (BMI), inadequate gestational weight gain (GWG), and significantly elevated levels of fasting blood glucose (FBG), oral glucose tolerance test (OGTT), glycated hemoglobin (HbA1c), homeostasis model assessment-insulin resistance (HOMA-IR), triglycerides (TG), and total testosterone (T) levels (all P < 0.05). Regarding placental blood flow, the umbilical artery pulsatility index (UAPI) was elevated and the cerebral-fetal ratio (CPR) was decreased in the FGR group (P < 0.05). Multivariate Logistic regression analysis revealed that GWG(OR=0.871, P=0.039), HOMA-IR (OR=1.421, P=0.037), anti-Müllerian hormone (AMH) (OR=1.263, P=0.015), UAPI (OR=4.873, P=0.005), and CPR (OR=0.112, P=0.002) were identi?ed as independent risk factors for FGR. The area under the curve (AUC) of the predictive model constructed based on ?ve variables was 0.892 (95%CI: 0.823–0.961), demonstrating good discriminatory ability. The Hosmer-Lemeshow test yielded a P value of 0.421, indicating good calibration. Conclusion In pregnant women with GDM and PCOS, inadequate GWG, metabolic abnormalities, endocrine disorders, and placental blood ?ow impairment are all significant risk factors for FGR. The predictive model developed based on these factors demonstrates high accuracy and clinical applicability, providing a reference for risk strati?cation and individualized management of high- risk pregnant women in late pregnancy.