The primary aim of this study was to assess the long-term clinical and radiographic outcomes of arthroscopic debridement and bone marrow stimulation for talar osteochondral defects. The secondary aim was to identify prognostic factors that affect the long-term results. Fifty (88%) of fifty-seven eligible patients with a primary osteochondral defect treated with arthroscopic debridement and bone marrow stimulation were evaluated after a mean follow-up of twelve years (range, eight to twenty years). Clinical assessment included the Ogilvie-Harris score, Berndt and Harty outcome question, American Orthopaedic Foot & Ankle Society (AOFAS) ankle-hindfoot score, and Short Form-36 (SF-36) as well as resumption of work and sports. Weight-bearing radiographs were compared with preoperative radiographs with use of an ankle osteoarthritis classification system. The size, location, and classification of the defect, patient age and body mass index, traumatic etiology, and duration of symptoms were recorded and analyzed with use of univariate logistic regression. The Ogilvie-Harris score was excellent in 20% of patients, good in 58%, fair in 22%, and poor in 0%. According to the Berndt and Harty outcome question, 74% of patients rated the ankle as good, 20% as fair, and 6% as poor. The median AOFAS score was 88 (range, 64 to 100). Of the eight subscales of the SF-36, six were comparable with population norms and two were superior in the study group. Ninety-four percent of patients had resumed work and 88% had resumed sports. The radiographs indicated an osteoarthritis grade of 0 in 33% of the patients, I in 63%, II in 4%, and III in 0%. Compared with the preoperative osteoarthritis classification, 67% of radiographs showed no progression and 33% showed progression by one grade. None of the prognostic factors was significantly associated with the Ogilvie-Harris score or progression of osteoarthritis. This study suggests that initial success of arthroscopic debridement and bone marrow stimulation for osteochondral defects of the talus are maintained over time. No factors that were predictive of the outcome could be identified.