Cell-based regenerative medicine, including tissue engineering, is a novel approach to reconstituting tissues that do not spontaneously heal, such as damaged cartilage, and to curing diseases caused by malfunctioning cells. Typically, manufacturing processes to generate cartilage for replacement therapies involve isolation and expansion of cells from cartilage biopsies. A challenge in the field is potential contamination by other cell types (e.g., fibroblast-like cells), which can overgrow the desired cells during culturing and may ultimately compromise clinical efficacy. No standard analytical system has been absolutely effective in ensuring the identity of these cell-based products. Therefore, we tested deoxyribonucleic acid methylation analysis as a quality assessment tool, applying it to Genzyme's Carticel product, a chondrocyte implant that the Food and Drug Administration has approved. We identified 7 potent discriminators by assaying candidate genomic regions derived from methylation discovery approaches and literature searches regarding a functional role of genes in chondrocyte biology. Using a support vector machine, we trained an optimal cell type classifier that was absolutely effective in discriminating chondrocytes from synovial membrane derived cells, the major potential contaminant of chondrocyte cultures. The abundant marker availability and high quality of this assay format also suggest it as a potential quality control test for other cell types grown or manipulated in vitro.