Purpose: Evaluation by animal models is essential for tissue engineering-based articular cartilage repair techniques. Larger animals are considered to more closely approximate the clinical situations in translational medicine. Therefore, we used miniature pigs and induced full-thickness and osteochondral defects in them. We also studied if there were instances of spontaneous repair for providing baseline data for further cartilage regeneration study. Methods: A total of 12 miniature pigs with average age of 7.4 months were used in this study. Full-thickness and osteochondral defects with 2.7, 4.5 or 8.0 mm diameter were created at medial femoral condyles in the same pig, respectively. The pigs were sacrificed at 8, 16, 24 and 48 weeks. Gross appearances of defects were observed at aforementioned time points, and the histological analyses, including H&E and alcian blue staining, were performed for consequent evaluation as well. Results: The results showed that defects created in the center of the femoral condoyle migrated to periphery, and it implied that the pigs were still growing after 7–8 months of age. However, spontaneous repair was observed in 2.7 mm diameter defects but rarely seen in 4.5 and 8 mm diameter osteochondral defects. On the other hand, osteochondral defects repaired better than defects of full-thickness in the same 2.7 mm diameter. Conclusions: In order to prevent spontaneous repair of osteochondral defect in a young miniature pig animal model, a critical-sized osteochondral defect larger than 40% width of femoral medial condyle (4.5 mm in miniature pig) and observation period for more than 48 weeks are suggested by this study.