The purpose of this study is to evaluate the influence of the extent of involvement in the results of cementless THA in patients with idiopathic AVN of the femoral head. We reviewed 70 hips (52 patients) who had undergone cementless THA for idiopathic AVN of the femoral head. Anatomic femoral components were used in 43 hips and Harris Galante porous prosthesis were employed in 27 cases. The hips were classified according to International Classification (Association Research Circulation Osseous). The length of the follow-up period ranged from 3 to 5 years (average 53.5 months). Preoperative and postoperative clinical documentation and radiographs were evaluated. Statistical analysis was performed on the results of three analysis groups (Group A, comparison of results among Stage IIIA, IIIB, IIIC, and IV; Group B, between stage III and IV; Group C, between IIIA+IIIB and IIIC + IV). Of the 70 hips in the present study, there were 47 hips in stage III (IIIA, 15; IIIB, 19; IIIC, 13) and 23 hips in stage IV. The average postoperative Harris hip score at the time of study was 91.5 (Analysis Group C, p=.009). Femoral subsidence of more than 5 mm occurred in 5 hips (7.1%). Two hips were in Stage IIIC and three hips were in Stage IV (Group C, p=.024). The hips lower than -10 on Engh's index were present in one hip in Stage IIIC and three hips in Stage IV (Group C, p=.011). This study demonstrated that the clinical and radiological results of cementless THA were poor when the extent of involvement was greater or in the case of late stage subjects, particularly those above IIIC (p<05), with avascular necrosis of the femoral head.