Objective To investigate the predictive value of apolipoprotein B （ApoB） for intracranial arterial stenosis in patients with acute cerebral infarction （ACI）. Methods Clinical data of 391 patients with ACI in the Department of Neurology， Shandong Provincial Qianfoshan Hospital， from August 1， 2021 to December 1， 2023 were retrospectively collected. Patients were categorized according to the degree of intracranial artery stenosis into mild stenosis group （n=223）， moderate stenosis group （n=89）， and severe stenosis group （n=79）. Factors influencing the degree of intracranial arterial stenosis were analyzed using ordered and binary Logistic regression. The assessment of model predictive efficacy was performed by building model nomogram， plotting receiver operating characteristic （ROC） curve， calibration curve， clinical impact curve （CIC）， and decision curve analysis （DCA）. Net reclassification index （NRI） and integrated discrimination improvement index （IDI） were used for the comparison of different predictive models. Results There were statistically significant differences in the proportions of hypertension， diabetes mellitus， hyperlipidemia， as well as the levels of low-density lipoprotein cholesterol， total cholesterol， triglyceride， ApoA1， and ApoB among the patients in the mild， moderate， and severe stenosis groups （all P ＜ 0.05）. Ordered multifactorial Logistic regression analysis showed that ApoB was an independent risk factor for increased intracranial arterial stenosis in ACI patients ［OR=10.350， 95%CI （3.466，30.939），P ＜ 0.001］. Binary multifactorial Logistic regression analysis showed that ApoB was independently associated with moderate stenosis ［OR=8.236， 95%CI （2.378，28.528）， P＜ 0.001） and severe stenosis ［OR=7.702， 95%CI （2.122， 28.633）， P＜ 0.05］， respectively. The concomitant rates of diabetes mellitus and hyperlipidemia， low-density lipoprotein cholesterol， total cholesterol， triglyceride， and ApoB levels of patients in moderate and severe stenosis group were higher than those in mild stenosis group， and the differences were all statistically significant （all P＜0.001）. Binary multivariate Logistic regression analysis showed that ApoB was an independent risk factor for moderate and severe stenosis of intracranial arteries in patients with ACI ［OR=21.417， 95%CI （6.052，73.887）， P ＜ 0.001］. ApoB was introduced into the traditional model to plot the nomogram of the new model. ROC curve showed that the predictive value of the new model （AUC=0.702， sensitivity=47.60%， specificity=87.00%） was better than that of the traditional model （AUC=0.660， sensitivity=45.80%， specificity=82.10%）. Calibration curves suggested that the predictive calibration curves of the two predictive models were in good agreement with the standard curve. DCA and CIC curves indicated that the net clinical yield of the new model was better than that of the traditional model. NRI and IDI of the new model were improved compared to the traditional model （NRI=0.455， P＜ 0.001； IDI=0.060， P＜ 0.001）. Conclusions ApoB may serve as a predictor of intracranial arterial stenosis in patients with ACI and provide a basis for early disease assessment in patients with ACI.