The purpose of this study is to evaluate the results of the reconstruction of acetabular bone defect with morselized bone graft in revision of failed acetabular cup, used screw fixed hemispherical cup or bipolar cup as a component. Forty five revisions of an acetabular component that had been performed in 42 patients between 1988 and 1993 at SNUH, and had followed more than two years were included in this study. The revised acetabulum was evaluated by Harris hip score, in terms of osseous union and resorption of the graft, the amount of the migration of the revised components. Cup angle change and radiolucency were also evaluated in screw fixed hemispherical cup. The mean Harris hip score was improved 54 to 87 in screw fixed hemispherical cup and 56 to 78 in bipolar cup. In screw fixed hemispherical cup, osseous union showed within 6 months is 91.3% (21 cases), the graft bone resorption less than 10% of the initial graft thickness is 73.9% (17 cases). In bipolar cup, osseous union within 6 months is 52.2% (12 cases), the graft bone resorption less than 10% of the initial graft thickness is 40.9% (9 cases). The failure of revised acetabular component was five in bipolar cup cases and one in screw fixed hemispherical cup cases. In conclusion, screw fixed hemispherical cup showed better clinical results, earlier graft osseous union, less graft bone resorption and lower failure rate than bipolar cup. There was no significant difference between pure allograft mixed with autograft in clinical and radiological results. So we recommend the technique used morselized allograft with screw fixed hemispherical cup, to reconstruct the acetabular bone defect in acetabular cup revision, though bipolar cup was used in staged operation in severe acetabular bone defect with was unable to fix the acetabular cup with screws.