Objective To evaluate the impact of "Free Medication Policy" in Tianjin on medication adherence among impoverished patients with severe mental disorders in the community. Methods A total of 14 935 impoverished patients with severe mental disorders in the 2021 Tianjin Mental Health Information System were selected as research subjects. The differences in medication adherence between 3 722 impoverished patients with severe mental disorders who participated in the "Free Medication Policy" and 11 213 impoverished patients with severe mental disorders who did not participate in the "Free Medication Policy" were compared， and the relevant factors of medication adherence were analyzed. Logistic multiple regression was used to analyze the impact of "Free Medication Policy" on patients' medication adherence. Results Among 14 935 impoverished patients with severe mental disorders， 3 722（ 24.92%） participated in the "Free Medication Policy". The proportion of patients who participated in the "Free Medication Policy" taking medication regularly was 95.43%（3 552/3 722）， which was significantly higher than the 87.75%（ 9 839/11 213） of patients who did not participate， and the difference was statistically significant（ χ2=178.09， P＜ 0.01）. The probability of patients participating in the "Free Medication Policy" taking medication regularly was 2.72 times that of patients who did not participate in the policy［ OR=2.72， 95%CI（ 2.29，3.23）］. Previous risky behavior［ OR=1.36， 95%CI（ 1.01， 1.84）］ and hospitalization［ OR=1.16， 95%CI（ 1.02，1.32）］ were protective factors for patients taking medication regularly， and schizophrenia［ OR=0.80， 95%CI（ 0.69，0.92）］ and weak monitoring［ OR=0.73， 95%CI（ 0.61， 0.87）］ were risk factors for patients taking medication regularly. Conclusions The "Free Medication Policy" improves the medication adherence of impoverished patients with severe mental disorders in the community. In community prevention and control work， special attention should be paid to the medication situation of patients with schizophrenia and weak monitoring， and policy promotion and guidance should be increased.