The purpose of the current study was to investigate the feasibility of the application of defect-size femoral implants in a rabbit model of established cartilage defects and compare this treatment to microfracturing. In 31 New Zealand White rabbits, a medial femoral condyle defect was created in each knee. After 4 weeks, 3 animals were killed for defect baseline values. In the other 28 rabbits, knees were sham-operated, treated with microfracturing, or treated by placing an oxidized zirconium (OxZr) or cobalt-chromium (CoCr) implant (theta articulating surface 3.5 mm; fixating pin of 9.1 mm length). These animals were sacrificed 4 weeks after treatment. Joints were evaluated macroscopically. Implant osseointegration was measured by automated histomorphometry, and cartilage repair was scored microscopically. Cartilage quality was analyzed macroscopically and microscopically. Bone-implant contact was 63.2% +/- 3.2% for CoCr and 62.5% +/- 3.2% for OxZr. Cartilage defects did not show complete healing, nor during subsequent sham-surgery or microfracturing. For all treatments, considerable cartilage damage in the articulating medial tibia, and degeneration of lateral tibial and femoral cartilage was observed (p < 0.05). Both CoCr and OxZr implant-treated defects showed an increase of cartilage degeneration compared to microfracturing and sham-operated defects (p < 0.05). Although only a single short-term follow-up period was investigated in this study, caution is warranted using small metal implants as a treatment for established localized cartilage defects because, even after 4 weeks in this model, the metal implants caused considerable degeneration of the articulating surface.