This study aims to systematically evaluate the efficacy of bone marrow mesenchymal stem cell-derived exosomes (BMSCs-Exo) in improving spinal cord injury (SCI) to mitigate the risk of translational discrepancies from animal experiments to clinical applications. We conducted a comprehensive literature search up to March 2024 using PubMed, Embase, Web of Science, and Scopus databases. Two researchers independently screened the literature, extracted data, and assessed the quality of the studies. Data analysis was performed using STATA16 software. A total of 30 studies were included. The results indicated that BMSCs-Exo significantly improved the BBB score in SCI rats (WMD = 3.47, 95% CI [3.31, 3.63]), inhibited the expression of the pro-inflammatory cytokine TNF-α (SMD = -3.12, 95% CI [-3.57, -2.67]), and promoted the expression of anti-inflammatory cytokines IL-10 (SMD = 2.76, 95% CI [1.88, 3.63]) and TGF-β (SMD = 3.89, 95% CI [3.02, 4.76]). Additionally, BMSCs-Exo significantly reduced apoptosis levels (SMD = -4.52, 95% CI [-5.14, -3.89]), promoted the expression of axonal regeneration markers NeuN cells/field (SMD = 3.54, 95% CI [2.65, 4.42]), NF200 (SMD = 4.88, 95% CI [3.70, 6.05]), and the number of Nissl bodies (SMD = 1.89, 95% CI [1.13, 2.65]), and decreased the expression of astrogliosis marker GFAP (SMD = -5.15, 95% CI [-6.47, -3.82]). The heterogeneity among studies was primarily due to variations in BMSCs-Exo transplantation doses, with efficacy increasing with higher doses. BMSCs-Exo significantly improved motor function in SCI rats by modulating inflammatory responses, reducing apoptosis, inhibiting astrogliosis, and promoting axonal regeneration. However, the presence of selection, performance, and detection biases in current animal experiments may undermine the quality of evidence in this study.