Aim: This retrospective study aimed at the statistical identification of risk factors for dislocation for a certain cementless hip endoprothesis. Material and method: At our department 2605 primary total hip arthroplasties were performed between 1987 and 1997. In 40 patients (1.5%) a dislocation occurred. These patients were compared with a control group matched in number. Results: No difference was found in both groups with respect to age, height, weight, body-mass index, and diagnosis. A tendency was found that more males were involved in the dislocation group. There were more previous operations found in this group (p = 0.005). Also significant was the fact, that patients in the dislocation group more often had epidural anaesthesia (p = 0.02), more often the implantation of the smaller 28-mm head ( vs. 32 mm) (p = 0.02), and a higher inclination angle of the acetabular component (p = 0.02). No difference was found in terms of the surgeons experience, the postoperative leg length, femoral offset, medialisation, cranialisation and antetorsion of the acetabular component. Conclusion: Dislocation after total hip arthroplasty is a multifactorial complication. With the use of the Alloclassic®/Zweymuller total hip endoprthesis, a higher risk for dislocation was found for patients with previous hip surgery, the use of the smaller endoprosthesis head, a high acetabular inclination angle, and in the use of epidural anaesthesia.