The results of forty core-decompression procedures that were performed for ischemic necrosis of the femoral head in thirty-one patients over a four-year period were retrospectively reviewed to ascertain the effectiveness of the procedure. The data did not support the published rates of success of the procedure for Stage-1 and Stage-2 lesions. With a mean length of follow-up of eighteen months, 60 per cent of the hips that had a decompression prior to collapse of the femoral head demonstrated progression of the lesion and were judged to be a failure by clinical or radiographic criteria. Computerized tomographic scans and magnetic resonance imaging proved to be 100 per cent sensitive for diagnosis in preoperative testing; isotopic bone-scanning was less sensitive (80 per cent). Functional intraoperative testing by the method of Ficat did not provide added sensitivity or specificity to the results of the preoperative bone scan. A postoperative or intraoperative fracture occurred in four hips, for an incidence that exceeded any in previously published reports. We concluded that core decompression should be considered a relatively ineffective procedure with significant morbidity.