We undertook a randomised controlled trial to compare bipolar hemiarthroplasty (HA) with a novel total hip replacement (THR) comprising a polycarbonate-urethane (PCU) acetabular component coupled with a large-diameter metal femoral head for the treatment of displaced fractures of the femoral neck in elderly patients. Functional outcome, assessed with the Harris hip score (HHS) at three months and then annually after surgery, was the primary endpoint. Rates of revision and complication were secondary endpoints. Based on a power analysis, 96 consecutive patients aged > 70 years were randomised to receive either HA (49) or a PCU-THR (47). The mean follow-up was 30.1 months (23 to 50) and 28.6 months (22 to 52) for the HA and the PCU group, respectively. The HHS showed no statistically significant difference between the groups at every follow-up. Higher pain was recorded in the PCU group at one and two years' follow-up (p = 0.006 and p = 0.019, respectively). In the HA group no revision was performed. In the PCU-THR group six patients underwent revision and one patient is currently awaiting re-operation. The three-year survival rate of the PCU-THR group was 0.841 (95% confidence interval 0.680 to 0.926). Based on our findings we do not recommend the use of the PCU acetabular component as part of the treatment of patients with fractures of the femoral neck.