PURPOSE: The goal of this study was to determine whether a biopsy specimen of a fixed size can be harvested reliably and consistently by arthroscopy at the posteromedial rim of the talar dome. METHODS: A cartilage biopsy specimen was taken post mortem arthroscopically from the posteromedial rim of the talar dome in 20 ankles. We aimed to take a full-thickness biopsy specimen of 10 x 5 mm in size. The shape, length, width, position, and depth of the created defect were determined. Subsequently, 2 observers analyzed the biopsy sites twice. The same set of parameters and the surface area of the lesion were determined. Differences between aimed and observed sizes were studied. RESULTS: In all ankles the aimed biopsy site could be seen and reached. The observed mean size of the biopsy specimens, when compared with the aimed size, was only significantly different for the mediolateral size. A wide variation of surface area was found. Nearly all biopsy specimens started exactly at the posterior border of the dome, but they were somewhat more lateral than intended. In only half of the biopsy specimens was the shape linear or oval, the others being rounded or irregular. In case a pre-existing posteromedial cartilage lesion was present, its position did not coincide with that of the biopsy specimen. CONCLUSIONS: A limited cartilage biopsy specimen could be harvested reliably and consistently by arthroscopy at the posteromedial rim of the talar dome, although the size tended to be somewhat larger than we intended. Containment of a pre-existing posteromedial lesion was never jeopardized. CLINICAL RELEVANCE: A safe biopsy site for cartilage in the ankle eliminates donor-site morbidity in the knee and may improve the suitability and biology of the obtained chondrocytes for successful repair of symptomatic joint surface defects of the ankle.