Abstract. To assess the clinical efficacy and safety of mesenchymal stem cell (MSC) treatment for osteoarthritis of the knee (KOA), a systematic electronic literature search was performed on PubMed, EMBASE and Web of Science. Studies published in English from the earliest record to December 2014 were searched using the following keywords: Cartilage defect, cartilage repair, osteoarthritis, KOA, stem cells, MSCs, bone marrow concentrate (BMC), adipose-derived mesenchymal stem cells, synovial-derived mesenchymal stem cells and peripheral blood-derived mesenchymal stem cells. The effect sizes of selected studies were determined by extracting pain scores from the visual analog scale and functional changes from International Knee Documentation Committee and Lysholm and Western Ontario and McMaster Universities Osteoarthritis Index before and after MSCs or reference treatments at 3, 6, 12, and 24 months. The factors were analyzed and the outcomes were modified after comparing the MSC group pooled values with the pretreatment baseline or between different treatment arms. A systematic search identified 18 clinical trials on this topic, including 10 single-arm prospective studies, four quasi-experimental studies and four randomized controlled trials that used BMCs to treat 565 patients with KOA in total. MSC treatment in patients with KOA showed continual efficacy for 24 months compared with their pretreatment condition. Effectiveness of MSCs was improved at 12 and 24 months post-treatment, compared with at 3 and 6 months. No dose-responsive association in the MSCs numbers was demonstrated. However, patients with arthroscopic debridement, activation agent or lower degrees of Kellgren-Lawrence grade achieved improved outcomes. MSC application ameliorated the overall outcomes of patients with KOA, including pain relief and functional improvement from basal evaluations, particularly at 12 and 24 months after follow-up.