The diagnostic accuracy of magnetic resonance imaging (MRI)-targeted biopsy by using the transperineal (TP) versus transrectal (TR) route in the detection of clinically significant prostate cancer (csPCa) remains to be revealed. A systematic search of PubMed, Embase, Ovid, and the Cochrane Library up to April 2019 was conducted. We pooled odds ratios with 95% confidence intervals (CIs) for csPCa detected by TP and TR MRI-targeted biopsy. The relative sensitivity (or risk ratio) between TP and TR route was synthesized. We also pooled the diagnostic sensitivity of either approach using the combined biopsy results as the reference standard. A total of 328 patients with positive multiparametric MRI underwent TP MRI-targeted biopsy, and 315 patients underwent TR MRI-targeted biopsy. The TP route detected more csPCa, with a detection rate of 62.2% (204/328) compared to 41.3% (130/315) for the TR route (odds ratio = 2.37; 95% CI, 1.71-3.26). After adjusting for differences in cancer prevalence, TP MRI-targeted biopsy detected 91.3% (105/115) of csPCa compared to 72.2% (83/115) by the TR route (risk ratio = 1.26; 95% CI, 1.02-1.54). The pooled diagnostic sensitivity of the TP route (86%; 95% CI, 77-96) was better than the TR route (73%, 62-88%). The TR approach missed more csPCa located at the anterior zone of the prostate (20 vs. 3). The TP route performed better than the TR route in MRI-targeted biopsy, especially in detecting csPCa located at the anterior prostate. More large prospective randomized or head-to-head comparison studies comparing the two approaches are warranted. Copyright © 2019 Elsevier Inc. All rights reserved.