Intertrochanteric osteotomy has been recommended as an alternative to total hip replacement for the treatment of osteoarthrosis of the hip in younger patients because of the higher risk of mechanical loosening after total hip replacement in this age-group. The results of 112 consecutive cemented total hip replacements that were performed after a medial-displacement intertrochanteric osteotomy had failed (the conversion group) were compared with the results of 262 consecutive primary total hip replacements with cement (the control group). The mean length of follow-up was fifty-six months for the first group and fifty-four months for the second. The mean time between the osteotomy and the total hip replacement was six years. Two hundred and four Muller prostheses and 170 Lubinus hip prostheses were used. Pain was satisfactorily relieved in 89 per cent of the group that had a conversion and in 91 per cent of the control group. There were only minor differences in the range of motion of the hip and in walking performance between the two groups. The femoral component was more frequently in a varus position in the control group (p less than 0.05). The radiographic findings, including migration of the stem and radiolucency around the stem, were not inferior in the group that had a conversion. The rate of intraoperative fracture was significantly greater in the group that had a conversion, and it was correlated with the amount of displacement of the osteotomy. There were four failures (1.5 per cent) in the control group and one (0.9 per cent) in the group that had a conversion.(ABSTRACT TRUNCATED AT 250 WORDS)