This study aimed to develop and validate a predictive model based on S100 calcium-binding protein B (S100B) and neuron-specific enolase (NSE) as the core of epilepsy secondary to cerebral infarction. For this aim, 156 cases of cerebral infarction from June 2018 to December 2019 were selected. According to the ratio of 7:3, 109 cases were used for training and 47 cases were used for validation. The factors influencing cerebral infarction secondary to epilepsy were analyzed by a univariate analysis comparing the general data of the two groups and binary logistic regression, and the prediction model was established and validated. Results showed that there was no statistically significant comparison of general information between the training and validation groups (p>0.05). The comparison of NIHSS score, lesion location, lesion size, infarct staging, involved arterial system, large infarct, NSE, and S100B levels between the two groups was significant (P<0.05). The difference between the two groups will be secondary epilepsy= 1, non-epilepsy=0 as dependent variables and factors with significant differences in the univariate analysis as covariates for logistic regression analysis showed that NIHSS score > 15, cortical lesion, lesion size ≥ 5cm, carotid circulation involvement, large infarct, S100B, NSE were risk factors for secondary epilepsy in cerebral infarction. In conclusion, serum S100B and NSE levels were abnormally elevated in patients with epilepsy secondary to cerebral infarction, NIHSS score > 15, cortical lesions, lesion size ≥ 5 cm, carotid circulation involvement, large infarct, S100B and NSE are risk factors for epilepsy secondary to cerebral infarction, and the AUC area of S100B and NSE is large, based on S100B and NSE as The prediction model based on S100B and NSE has good predictive value.