OBJECTIVE Total hip arthroplasty (THA) is generally proposed for renal transplant patients with invalidating hip disease. For patients on chronic dialysis, the few published series report a higher rate of complications. These patients are considered more vulnerable. We report the results of a retrospective mid-term analysis of 28 renal failure patients (37 hips) with THA comparing renal transplant recipients with chronic dialysis patients. METHODS THA was performed on 37 hips in 28 patients with renal failure between January 1993 and 2004. Treatment for the renal disease was transplantation or chronic dialysis. Mean patient age at the time of the arthroplasty was 56 years. Nine patients had bilateral THA. At review, four patients had died and none were lost to follow-up. Mean time between transplantation and hip arthroplasty was 9.9 years. Among the fourteen dialysis patients (21 hips), five had had a renal transplant. Mean time from onset of dialysis to arthroplasty was 12.3 years. At mean postoperative follow-up of six years (range 24 months to 12 years), 24 patients were living. Thirty-five hips had no surgical history. Aseptic osteonecrosis was the dominant etiology (75% of operative indications in the graft group and 52% in the dialysis group). The preoperative Postel-Merle-d'Aubigné score was 7.6 in the dialysis group and 10.3 in the graft group. Most of the cups were not cemented (n=31, 84%), as were most of the stems (n=29, 78%). All survivors were reviewed. The PMA score was determined. RESULTS Early postoperative complications in the dialysis patients were: deep vein thrombosis (n=3), pulmonary embolism (n=4), operative site hematoma (n=2) and immediate septic complication with prompt revision and preservation of the implants (n=1). For the graft group, complications were: early dislocation (n=2), lung disease (n=3). Four patients, all in the dialysis group, died. At review, the mean PMA score was 14.2 in the dialysis patients and 15.7 in the graft patients. There were no radiological signs of loosening, nor of polyethylene wear, and no ectopic ossifications could be identified. There were no late infections. CONCLUSIONS In renal transplant recipients, total hip arthroplasty is a reliable treatment for hip disease, providing good mid-term results and a morbidity close to that observed in the general population. Conversely, dialysis patients have a greater perioperative morbidity. Use of non-cemented implants is not associated with a higher rate of loosening than with cemented implants. It can be recalled that this type of surgery must be performed within the framework of careful pluridisciplinary patient management.