Purpose: The use of mesenchymal stem cells (MSCs) represents a promising alternative cellular model to chondrocytes for articular cartilage repair but satisfactory protocols allowing proper chondrogenic conversion together with sufficient cartilage matrix production are still missing. The purpose of this study was to undertake an extensive characterization of human MSCs isolated from Wharton’s jelly (WJ), adipose tissue (AT), dental pulp (DP) and bone marrow (BM), and to confront their chondrogenic potential in serum-free culture conditions. Methods: MSCs were isolated and expanded in serum-free conditions. Their proliferation kinetics were evaluated. An extensive immunophenotyping analysis by flow cytometry was undertaken at the end of their first passage (P1). We then explored their potential to engage into adipogenic, osteogenic and chondrogenic lineages, the latter being assessed in serum-free conditions with a view of clinical application. This chondrogenic conversion was performed in micromass pellets and in another cellular model using MSCs embedded in agarose hydrogel. This model offers two advantages: 1) it mimicks a cartilage tissue engineering protocol where MSCs converted into chondrocytes and embedded in hydrogel form a cartilage gel that could be used to treat cartilage lesions and 2) the agarose nature of the hydrogel allows the release of the cells from the gel (by enzymatic digestion) at the end of their chondrogenic conversion, giving therefore the possibility to analyze them directly by flow cytometry. Recently, we developed and characterized the first antibody capable of detecting the IIB isoform of human type II procollagen, the only isoform of this collagen type that is expressed by well-differentiated chondrocytes. Here, we used this antibody by flow cytometry to look for intracellular IIB procollagen expression in MSCs undergoing chondrocyte differentiation in hydrogel. Results: The mean population doubling times between P1 and P5 ranged between 30 h and 55 h, depending on the tissue origin of MSCs. The doubling times remained constant from P1 to P5 and a minimal reservoir of 300 × 10ˆ6 cells (e.g., BM-MSCs) could be obtained after 5 passages, starting with 1 × 10ˆ6 cells harvested at the end of P1. Importantly, no karyotype abnormalities were detected after 5 passages. In brief, all cell sources yielded clinical-scale amounts of MSCs after expansion in serum-free culture conditions. By using an original panel of 27 surface markers that were analyzed by 8-color flow cytometry, our results revealed that all MSC sources were positive for the expression of the ISCT-validated markers (CD73, CD90, CD105) as well as for other recognized MSC markers (CD44). Nevertheless, differences in expression for other markers were found between the sources. BM-MSCs were the only cells to include subpopulations positive for all the putative skeletal markers analyzed (CD29, CD56, MSCA-1, Stro-1, CD106, CD146, CD271). All cell sources showed signs of adipogenic, osteogenic and chondrogenic conversion in monolayer cultures, after amplification in serum-free conditions. However, only BM-MSCs were able to clearly convert into true chondrocytes in micromass pellets or hydrogel cultures, as attested in particular by the expression of type IIB procollagen, monitored by immunohistochemistry or flow cytometry. Conclusions: It is possible to drive MSCs to the chondrogenic lineage in hydrogel after expansion and differentiation in serum-free conditions. BM-MSCs appear to be the best candidates for chondrogenic conversion in the conditions used in this study and this is in concordance with the largest panel of putative skeletal markers revealed by flow cytometry analysis for this category of MSCs. In this context, we have introduced an innovative quality control of the chondrocytic status of MSCs embedded in hydrogel, via measurement by flow cytometry of intracellular IIB procollagen, a marker of true chondrocytes.