BACKGROUND: Haloperidol remains a frequently prescribed first-generation antipsychotic. However, haloperidol-associated mortality risk by all causes, cardiovascular disease (CVD), and pneumonia compared with other antipsychotics is unknown. OBJECTIVE: This study investigated the mortality risk associated with long-term haloperidol treatment versus that with other antipsychotics. METHODS: We identified incident antipsychotic users from 2004 to 2014 in the Clinical Data Analysis and Reporting System (CDARS), a population-based clinical database managed by the Hong Kong Hospital Authority. We included patients who were aged ≥ 18 and received antipsychotics for any indication apart from terminal illnesses or management of acute behavioural disturbance. Patients on haloperidol and other antipsychotic agents (risperidone, quetiapine, olanzapine, chlorpromazine, aripiprazole, sulpiride, amisulpride, or trifluoperazine) were matched by propensity score. Hazard ratios (HRs) for all-cause mortality and death due to CVD and pneumonia were estimated with 95% confidence intervals (CIs) using a Cox proportional hazards model. RESULTS: In total, 136,593 users of antipsychotics were included. During a mean follow-up of 3.2 years, the incidence of all-cause mortality ranged from 186.8/1000 person-years for haloperidol to 10.4/1000 person-years for trifluoperazine. The risk of all-cause mortality was lower with non-haloperidol antipsychotics than with haloperidol, with HRs ranging from 0.68 (95% CI 0.64-0.72 [chlorpromazine]) to 0.43 (95% CI 0.36-0.53 [trifluoperazine]). Risperidone, quetiapine, sulpiride, chlorpromazine, aripiprazole, and trifluoperazine were associated with a significantly lower risk of pneumonia-related mortality. A significantly lower risk of CVD mortality was observed for risperidone, sulpiride, chlorpromazine, and quetiapine. CONCLUSION: Haloperidol was associated with increased overall mortality when compared with other antipsychotics in long-term follow-up. Treatment with haloperidol should be carefully considered, especially in older patients and patients at risk of CVD or pneumonia, since the risk of death appears to be lower with non-haloperidol agents.