Objective: To evaluate the effects and safety of intra-articular injection of mesenchymal stem cells on patients with knee osteoarthritis by a systematic review and meta-analysis. Methods: PubMed, EMBASE, and Cochrane Library were retrieved. An assessment of the risk of bias was done through the Cochrane Collaborative Bias Risk Tool, publication bias was assessed by plotting funnel plots and Egger tests. Pain and functional improvements in patients with knee osteoarthritis were determined by changes in VAS scores and WOMAC scores at baseline and follow-up endpoints. For the evaluation of MRI, the WORMS score and changes in cartilage volume were used. In addition, the number of adverse events in the intervention group and the control group were counted to explore the safety. Results: A total of 10 randomized controlled trials involving 335 patients were included. In the pooled analysis, compared with the control groups, the VAS scores of MSC groups decreased significantly (MD,−19.24; 95% CI: −26.31 to −12.18, P < .00001. All of the WOMAC scores also improved significantly: the total scores (SMD, − 0.66; 95% CI: − 1.09 to −0.23, P = .003), pain scores (SMD, − 0.46; 95% CI: − 0.75 to −0.17, P = .002), stiffness scores (SMD, −0.32; 95% CI: −0.64 to 0.00 P = 0.05), and functional scores (SMD, −0.36; 95% CI: −0.69 to −0.04, P = .03). Two studies with non-double-blind designs were the main source of heterogeneity. In terms of cartilage repair, there was no significant difference in the WORMS score, but there was a significant increase in cartilage volume in the MSC group (SMD, 0.69; 95% CI: 0.25 to 1.13, P = .002). The proportion of patients with adverse events in the MSCs treatment group was significantly higher than that in the control group (OR, 3.20; 95% CI: 1.50 to 6.83, P = .003). Conclusions: Intra-articular injection of mesenchymal stem cells is effective and safety to relieve pain and improve motor function of patients with knee osteoarthritis in a short term which is different to conclusions of previous study.