BACKGROUND: Case reports implicate disulfiram treatment in causing parkinsonism, but these observations lack epidemiological confirmation. Aim of the present study was to estimate the risk of incident parkinsonism associated with disulfiram dispensing in a large Italian population. METHODS: In this observational cohort study, administrative data were used, linking records at the individual level from civic registries, population census, mortality registers, hospital admissions, archives of drug prescriptions, and direct ambulatory drug distribution. Participants included all residents in the Piedmont region of Italy aged ≥ 40 years participating in 2011 census, still resident and alive at the beginning of 2013, followed-up from 2013 to 2019. The outcome was incident parkinsonism identified through multiple prescriptions of levodopa or a hospital admission for Parkinson's disease or atypical parkinsonism. Exposure to disulfiram and to neuroleptics was assessed through regional drug prescription archives. The association between disulfiram and parkinsonism onset was assessed using Cox proportional hazards models, adjusted for gender, age and neuroleptic use. RESULTS: The study population included 2,498,491 individuals (mean age: 62 years). During follow-up, 19,072 parkinsonism cases were identified, 8 of whom had been prescribed disulfiram. Exposure to disulfiram was associated with a three-fold increased risk of parkinsonism (HR = 3.10, 95% CI = 1.55-6.21) that remained significant when adjusted for neuroleptic use (HR = 2.04, 95% CI = 1.01-4.10). The association was stronger among persons unexposed to neuroleptics and among those with more than four disulfiram prescriptions. CONCLUSIONS: These results support the hypothesis that disulfiram may cause parkinsonism. Clinicians and drug regulatory agencies should consider parkinsonism when assessing the risks and benefits of disulfiram use.