BACKGROUND: Our study was conducted to investigate the effect of 5-ARI on the death rate from prostate cancer by means of an updated meta-analysis using related data. METHODS: We did comprehensive literature searches using the PubMed, Embase, and Cochrane databases through July 2019. We evaluated the risk of bias in such studies using the ROBINS-I tool and analyzed deaths from cancer and deaths from all causes using HR. RESULTS: The meta-analysis included 11 studies. The pooled overall HRs for cancer-specific deaths between 5-ARI treatment versus non-exposed groups were 0.937 (95% CI: 0.730, 1.201). In the subgroup moderator analysis, the number of patients and additional medication variables were significantly associated with deaths from all causes (p = 0.022 and p = 0.005, respectively). For detecting the publication bias or small-study effect in the included studies, we performed Begg and Mazumdar's correlation test (0.851 and 0.573) and Egger's regression coefficient test (0.035 and 0.245) and suggested that there was no evidence of publication bias or small-study effect in this meta-analysis. CONCLUSIONS: Our study indicated that exposure to 5-ARI had no close association between the overall death rate or cancer-specific deaths. To confirm these results, well-designed prospective studies with large samples are required.