Osteochondral injury is therapeutically irreversible within current treatment parameters. Autologous chondrocyte implantation ( ACI) promises to regenerate hyaline articular cartilage, but conventional ACI is plagued by complications determined by periosteal grafting. Here we propose the utilization of collagen membrane in ACI as an effective bioscaffold for the regeneration of osteochondral lesions. Using a rabbit model of osteochondral injury, we have inoculated autologous chondrocytes onto a type I/ III collagen scaffold [ so- called matrix- induced ACI ( MACI)] and implanted into 3- mm osteochondral knee defects. All untreated defect histology showed inferior fibrocartilage and/ or fibrous tissue repair. In our time- course study, ACI with type I/ III collagen membrane regenerated cartilage with healthy osteochondral architecture in osteochondral defects at 6 weeks. At 12 weeks, articular cartilage regeneration was maintained, with reduced thickness and proteoglycan compared with the adjacent cartilage. Both 6- week ( p < 0.01) and 12- week ( p < 0.05) ACI with collagen membrane showed significant improvement as compared with untreated controls. To further examine the efficacy of cartilage regeneration by ACI, we conducted a dose - response study, using chondrocytes at various cell densities between 10(4) and 10(6) cells/ cm(2). The results showed that cell density had no effect on outcome histology, but all cell densities were significantly better than untreated controls ( p < 0.01) and cell- free collagen membrane treatment ( p < 0.05). In short, our data suggest that autologous chondrocyte- seeded type I/ III collagen membrane is an effective method for the treatment of focal osteochondral knee injury in rabbits.