ObjectivesSchizophrenia is a chronic and severe mental disorder affecting more than 21 million people worldwide. Short-term exposure to nitrogen dioxide (NO2) has been associated with hospital admissions (HAs) for mental disorders, but no study has evaluated the specific association of NO2 and schizophrenia. Additionally, the shape of the concentration–response (C–R) curve has not yet been assessed at present. This study aims to investigate the relationship between short-term exposure to NO2 and HAs for schizophrenia in Hefei, from 2014 to 2016. We also attempt to explore the C–R and the underlying effect modifiers of the association.MethodsDaily number of HAs for schizophrenia was derived from the computerised medical record system of Anhui Mental Health Center. We used a time-series Poisson generalised linear regression combined with distributed lag non-linear models to model the NO2–schizophrenia relationship.ResultsA total of 11 373 HAs were identified during the study period. An increase in levels of NO2 was significantly associated with elevated schizophrenia HAs. The estimated relative risk per IQR increase in NO2 at lag 01 was 1.10 (95% CI 1.01 to 1.18). Greater association was observed in young patients (relative risk: 1.11, 95% CI 1.02 to 1.19). The modelled C–R curves of the NO2–schizophrenia relationship suggested possible threshold effects of NO2 for all ages combined, young patients, men and both seasons.ConclusionsShort-term exposure to NO2 may be associated with increased schizophrenia HAs. Findings indicated potential threshold effects of NO2, which has important implications for health-based risk assessments.