Evidence on acute respiratory failure (ARF) from antipsychotics is scant, and only 1 population-based study examined this drug safety issue in chronic obstructive pulmonary disease patients. Antipsychotics have been frequently prescribed off-label in adults, but whether antipsychotic use carries an increased ARF risk among adult patients is uncertain. Methods: We adopted a nested case-control study analysing 716 493 adults aged ≥20 years, identified from the Taiwan nationwide healthcare claims records between January 2000 and December 2013. Among the study cohort, 7084 adults with ARF and 12,785 disease risk scored-matched randomly selected controls were analysed. Multivariable logistic regression models were employed to estimate odds ratios of ARF with antipsychotic usages. Results: Current, recent, and recent past use of antipsychotics was associated with a 2.33-fold (95% confidence interval [CI] = 2.06-2.64), 1.79-fold (95% CI = 1.43-2.25) and 1.41-fold (95% CI = 1.20-1.66) increased risk of ARF, respectively, compared with nonuse, while antipsychotics discontinued >90 days carried no risk. A dosedependent association was observed with current therapy of antipsychotics (test for trend, P < .001), in which antipsychotic use at >1 defined daily dose yielded the highest risk of 6.53-fold (95% CI = 3.33-12.79). The findings were robust to using carbamazepine as an active comparator. Conclusion: Antipsychotic use was associated with an increased risk of ARF in adult patients. The risk was dose-dependent and markedly higher with current use of antipsychotic agents at doses of 1 defined daily dose and above, <10% of this cohort. Physicians should be vigilant about any respiratory symptoms in patients currently receiving antipsychotics at such dose.