Antipsychotics (AP) are widely used to treat schizophrenia and other psychiatric disorders. However, the association between the AP use and mortality risk is controversial. We searched PubMed, EMBASE, MEDLINE, PsycINFO, CINAHL, the Cochrane Library and four Chinese databases from inception to June 2016. All observational cohort or case–control studies reporting data on mortality outcomes in individuals exposed to AP drugs were included. This systematic review included 68 studies involving 4,812,370 participants. Sixty-seven studies reported confounding factors, the most common being age, sex, race, concomitant medications, and comorbidities. For all-cause mortality, current users of AP and conventional antipsychotics (CAP) had higher mortality risk than did non-AP users [AP users: RR, 1.50; 95% CI, 1.12 to 1.99; CAP users: RR, 1.53; 95% CI, 1.16 to 2.04]. However, the association between the current use of atypical antipsychotics (AAP) and the mortality was of borderline significance, and there was no significant difference for past users of AP. Mortality was higher in current CAP users than in current AAP users. For cardiac death and sudden death, current AP and CAP users also had higher mortality risk than non-AP users. A subgroup analysis showed a possible increased risk in patients with Parkinson’s, but not in those with dementia, Alzheimer’s disease, schizophrenia, delirium or stroke. An increased risk of all-cause mortality for patients ≧65 years may also exist. AP exposure is associated with an approximately 1.5-fold increased mortality risk. This increased risk may be particularly prominent in patients with Parkinson’s and those over 65 years old. Further studies are required to evaluate the mortality risk for individual AP drugs and diseases.