mbryonic stem (ES) cells are a potential source for cartilage tissue engineering because they provide an unlimited supply of cells that can be differentiated into chondrocytes. So far, chondrogenic differentiation of both mouse and human ES cells has only been demonstrated in two-dimensional cultures, in pellet cultures, in a hydrogel, or on thin biomaterials. The next challenge will be to form cartilage on a load-bearing, clinically relevant–sized scaffold in vitro and in vivo, to regenerate defects in patients suffering from articular cartilage disorders. For a successful implant, cells have to be seeded efficiently and homogenously throughout the scaffold. Parameters investigated were the scaffold architecture, seeding method, and cellular condition. Seeding in a three-dimensional fiber-deposited (3DF) scaffold was more homogenous than in a compression-molded scaffold. The seeding efficiency on bare scaffolds was compromised by the absence of serum in the chondrogenic medium, but could be improved by combining the cells with a gel and subsequent injection into the 3DF scaffolds. However, the viability of the cells was unsatisfactory in the interior of the graft. Cell aggregates, the so-called embryoid bodies (EBs), were seeded with increased survival rate. Mouse ES cells readily underwent chondrogenic differentiation in vitro in pellets, on bare scaffolds, in Matrigel, and in agarose, both as single cells and in EBs. The differentiation protocol requires further improvement to achieve homogenous differentiation and abolish teratoma formation in vivo. We conclude that ES cells can be used as a cell source for cartilage tissue engineering, pending further optimization of the strategy.