The purpose of this study was to investigate the short-term fate of arthroscopically implanted 6.0mm osteochondral allografts. The experiment was divided into a preliminary and experimental phase. During the preliminary phase techniques were developed to aseptically harvest osteochondral grafts from fresh bovine metacarpophalangeal (fetlock) joints, and an arthroscopic approach to the same joint was developed. In the experimental phase, osteochondral dowels (6.0mm in diameter and 10.0 to 15.0mm in length) and bone dowel controls were harvested from the lateral and medial condyle of cadaver fetlock joints of calves weighing 200kg. These grafts were refrigerated for 12 hours, and then inserted into 6.0mm drill holes in the lateral and medial condyle of both front fetlock joints in six calves using arthroscopic techniques. One fetlock joint of each calf received two grafts with intact cartilage, while the contralateral control received two bone dowels as negative controls. In the six recipient calves, a total of 24 grafts (12 osteochondral and 12 bone grafts) were implanted arthroscopically. The calves were recovered from anesthesia and allowed full activity during the three month period before they were slaughtered. Paravital staining, chondrocyte counts, and a modified Mankin scoring system were used to evaluate the joint surface in and around the experimental sites. Histology and measurement of the trabecular bone area under the grafts were used to assess the incorporation of the bony portion of the grafts. The results were compared to eight unoperated front fetlock joints of calves weighing 200kg. Overall, osteochondral grafts maintained an articular surface that was inferior to unoperated animals, but were far superior to bone dowel controls. Osteochondral allografts had good bony incorporation into the recipient sites, but postoperative congruency with the surrounding articular surface was essential for a good outcome. Five of 12 osteochondral grafts that were recessed or otherwise malpositioned became covered with fibrous tissue. Resurfacing of chondral defects with multiple small osteochondral grafts (MosaicPlasty) is currently used in human surgery, and may be useful in selected cases in animals. The results of this study show that long-term stability and congruency are essential, particularly in joints were the cartilage is thin. In areas where cartilage is thin, small mistakes result in exposure of the bony sidewall of the recipient site. New connective tissue develops from the exposed bone and spreads across the cartilage, ultimately destroying the graft.