Mesenchymal stem cells (MSCs) have emerged as a promising option to treat articular defects and early osteoarthritis (OA) stages. However, both their potential and limitations for a clinical use remain controversial. Thus, the aim of this systematic review was to examine MSCs treatment strategies in clinical settings, in order to summarize the current evidence of their efficacy for the treatment of cartilage lesions and OA.Among the 60 selected studies, 7 were randomized, 13 comparative, 31 case series, and 9 case reports; 26 studies reported the results after injective administration, whereas 33 used surgical implantation. One study compared the two different modalities. With regard to the cell source, 20 studies concerned BMSCs, 17 ADSCs, 16 BMC, 5 PBSCs, 1 SDSCs, and 1 compared BMC versus PBSCs. Overall, despite the increasing literature on this topic, the evidence is still limited, in particular for high-level studies. On the other hand, the available studies allow to draw some indications. First, no major adverse events related to the treatment or to the cell harvest have been reported. Second, a clinical benefit of using MSCs therapies has been reported in most of the studies, regardless of cell source, indication, or administration method. This effectiveness has been reflected by clinical improvements and also positive MRI and macroscopic findings, whereas histologic features gave more controversial results among different studies. Third, young age, lower BMI, smaller lesion size for focal lesions, and earlier stages of OA joints have been shown to correlate with better outcomes, even though the available data strength does not allow to define clear cutoff values. Finally, definite trends can be observed with regard to the delivery method: currently cultured cells are mostly being administered by i.a. injection, while one-step surgical implantation is preferred for cell concentrates. In conclusion, while promising results have been shown, the potential of these treatments should be confirmed by reliable clinical data through double-blind, controlled, prospective and multicenter studies with longer follow-up, and specific studies should be designed to identify the best cell sources, manipulation, and delivery techniques, as well as pathology and disease phase indications.