Objective: To investigate the influencing factors of postoperative infection in cesarean section patients and construct a nomogram prediction model. Methods: The clinical data of patients undergoing cesarean section in Shanxi Bethune Hospital from January 2021 to December 2021 were retroanalyzed and devided into infection group (n=60) and no infection group (n=1 086). The influencing factors of postoperative infection in cesarean section patients were selected into the nomogram risk prediction model. The discrimination, calibration, and prediction efficacy of the nomograms prediction model were evaluated. Results: The results of the multivariate Logistic regression analysis showed that having stable work was a protective factor for postoperative infection in cesarean section patients (OR=0.570, 95%CI: 0.331-0.983, P=0.043). Hypertensive disorders in pregnancy (OR=2.356, 95%CI: 1.324-4.192, P=0.004), GBS colonization (OR=3.154, 95%CI: 1.118-8.897, P=0.030), times of vaginal examinations >5 (OR=2.470, 95%CI: 1.146-5.324, P=0.021) and oxytocin induction (OR=2.457, 95%CI: 1.230-4.907, P=0.011) were independent risk factors of postoperative infection in cesarean section patients. The nomogram prediction model was established based on the above factors, whose consistency index (C-index) was 0.721 (95%CI: 0.651-0.791), and 0.051 was selected as the cut-off value, with a sensitivity of 66.7% and a specificity of 72.5%. The Homer-Lemeshow goodness-of-fit test suggested that the prediction model had a better calibration ability ( χ2=2.169, P=0.825). Conclusions: The nomogram prediction model constructed based on work, gestational hypertensive diseases, GBS colonization, times of vaginal examinations, and oxytocin labor induction has good accuracy and differentiation. It can identify the high-risk group after cesarean section to be infected in the early stage according to the predicted risk value, and corresponding preventive measures should be taken to reduce the occurrence of infection. However, further external verification and prospective comparative trials are needed to confirm the reliability of the predictive ability of the model.