Objective To investigate the factors related to postoperative cognitive impairment (CI) in patients with intracerebral hemorrhage (ICH) and construct a risk prediction model. Methods A total of 72 patients with ICH in General Intensive Care Unit, Zhengzhou Sixth People’s Hospital, Zhengzhou, from November 2022 to November 2024 were selected and divided into CI group (29 cases) and non-CI group (43 cases) according to whether CI occurred. The clinical data of the two groups were compared, and the prediction model of postoperative CI in ICH patients was established by Logistic regression. ROC analysis was used to evaluate the predictive efficiency of the model. Results Compared with the non-CI group, the CI group had a higher proportion of lobular lesions and multiple lesions, a larger amount of intracranial hematoma, lower intraoperative MAP, more patients admitted to ICU, a higher proportion of secondary epilepsy, and a higher discharge NIHSS score (P<0.05). Logistc regression analysis showed that lesion location (OR=2.081, 95%CI: 3.824~9.532), intracranial hematoma volume (OR=1.325, 95%CI: 1.121~5.718), secondary epilepsy (OR=1.528, 95%CI: 1.124~4.271) and NIHSS score (OR=2.144, 95%CI: 1.272~7.262) were the main influencing factors for postoperative CI in ICH patients (P<0.05). The CI risk prediction model for ICH patients after surgery was constructed according to the Logistic regression analysis results: Logit(P)=1.325+1.131×lobular location+0.842×intracranial hematoma+1.252×secondary epilepsy+1.582×discharge NIHSS score. ROC analysis showed that the AUC value of the CI risk prediction model for ICH patients was 0.794 (95%CI: 0.682~0.907, P<0.05), with a sensitivity of 82.72% and a specificity of 77.86%. Conclusion Lesion location, intracranial hematoma amount, continued epilepsy, and discharge NIHSS score are the main influencing factors for postoperative CI in ICH patients. Therefore, a multi-factor prediction model can be built to assess the risk of CI in ICH patients and guide follow-up treatment.