Objective To construct and validate a risk prediction model for medication discrepancy in the hospital-to-home transition for stroke patients, and to develop a simple risk assessment scale, so as to provide a reference for standardizing medication discrepancy management to ensure safety and continuity of patient medication. Methods In the cerebrovascular disease data platform of a tertiary hospital in Bengbu, 400 stroke patients discharged for 2 weeks were selected as a modeling group, and the risk prediction model of medication discrepancy was constructed by using univariate and logistic regression to analyze its influencing factors. The Hosmer-Lemeshow test and receiver operating characteristic analysis were performed to assess the accuracy of fit and prediction value of the model. Homophily sampling of 210 patients matched on basic data was performed for external validation of the model, and a simple risk assessment scale for medication discrepancy was developed. Results Medication discrepancy occurred in 245 cases (61.25%) in the modeling group. The results of logistic regression analysis showed that, caregiver, completeness of medication history, outpatient emergency or hospitalization experience within six months, number of long-term medications, taking high-risk medications, and medication adherence were the independent influencing factors of medication discrepancy in stroke patients (all P＜0.05). The Hosmer-Lemeshow test of fit P=0.814, area under the ROC curve (AUC) was 0.929, sensitivity was 0.873 and specificity was 0.826; in the validation group, the AUC was 0.925 and the correct prediction rate was 88.57%. A simple risk assessment scale with a total score of 10, a cutoff value of 4, and an AUC of 0.913 was finally developed, and the correct prediction rate of the scale was 85.24%. Conclusion The constructed risk prediction model for medication discrepancy in the hospital-to-home transition period of stroke patients has high sensitivity and specificity, it can effectively predict the occurrence of medication discrepancy. The simple risk assessment scale is a simple and feasible quantitative tool that can be used by clinical staff to measure patients at high risk of medication discrepancy, in order to take preventive measures to avoid medication discrepancy.