Introduction. The long-term wear performance of highly cross-linked polyethylene (HXLPE) in cemented total hip arthroplasty (THA) has rarely been reported. Here we report a prospective randomized comparative analysis of radiographic wear after a minimum follow-up of 10 years in cemented THAs with either HXLPE or conventional polyethylene (CPE), and assess its clinical relevance. Patients and Methods. From 1999 to 2001, we conducted 94 primary cemented THAs with a 22.225-mm head at our hospital as part of a prospective randomized trial. All surgeries were performed using a direct lateral approach with a trochanteric osteotomy (Dall's approach). The patients were divided into 4 groups. Twenty-six hips in group A were implanted with CPE sockets against zirconia heads and Charnley-type stems. HXLPE sockets (Aeonian, Kyocera Medical Corp) were implanted in all hips in the other 3 groups. Twenty-five hips in group B were implanted with zirconia heads and KC stems (Kyocera Medical Corp), 23 hips in group C with zirconia heads and distal cylindrical stems, and 20 hips in group D with stainless steel heads and C-stem (DePuy Inc). The sockets were highly cross-linked by gamma irradiation at a dose of 35 kGy, heat annealed at 110ºC, and sterilized with 25 kGy of gamma irradiation in nitrogen. For radiographic evaluation, anteroposterior radiograms were taken for each patient annually, and every two years postoperatively for wear analyses. Two-dimensional head penetration was measured on each postoperative radiogram using a computer-aided technique. Results. Wear measurements were performed for 59 cases followed up over 10 years. Linear wear rates were 0.138±0.074 (mm/year±SD) for group A, 0.010±0.015 for group B, 0.013±0.020 for group C, and 0.012±0.027 for group D. Linear wear rates differed significantly between group A and other groups, and no significant difference was found among groups B, C, and D. There were four revision cases. Among them, two sockets of group A were revised for aseptic loosening at 7 and 14 years postoperatively with linear wear rates of 0.749 and 0.153 mm, respectively. Two stems of group B and C were revised for aseptic loosening at 10 and 9 years postoperatively with linear wear rates of 0.007 and 0.041 mm, respectively. There were no other cases with aseptic loosening in any group. Osteolysis was found in 10 cases (group A: 7, group B: 1, group C: 1, group D: 1), and there was a significant difference in linear wear rates between the cases with and without osteolysis (0.157±0.083 and 0.030±0.053 mm/year±SD respectively). Discussion. The two revision cases of HXLPE did not have aggressive socket wear, and possibly cement fracture caused osteolysis and stem instability. The results of this study indicate that there is a significant difference in wear rate between CPE and HXLPE, and it was evident that PE wear was associated with osteolysis and aseptic loosening of the socket