Purpose of the study The longevity of an acetabular implant depends in part on its resistance to mechanical stress. Long-term stability of an uncemented cup requires stable primary fixation. This can be achieved with press-fit components. Material and methods One hundred forty-nine Duroloc® press-fit acetabular implants were studied after a minimum follow-up of 60 months. Among these 149 cups, 100 were analyzed with a mean follow-up of 60 months. Twenty-two patients were lost to follow-up, 19 were contacted by telephone, and eight had died. Clinical and radiological criteria were used to assess outcome. Digitalized x-rays were processed with Imagika (View Tec, Saint-Maurice, France) to determine the position of the implant in two-dimensional space, inclinaison, anteversion, polyethylene wear, bone-implant contact, and presence of ossifications. Seventy percent of the patients underwent arthroplasty for degenerative hip disease. Results The cohort included 40 women and 60 men. Mean age at surgery was 59.4 years, range 30-80 years. The Postel-Merle-d’Aubigné score improved from 10 to 16. Mean polyethylene wear was 0.5 mm with an annual rate of 0.087 mm/yr. Among the 57 cup-bone spaces present postoperatively, 41 had filled at last follow-up (78%). There were no cases of osteolysis. Peri-prosthetic ossifications were observed in 50 patients (16 stage I, 12 stage II, 19 stage II, and 3 stage IV). Medial displacement of the cup > 4 mm was observed in one patient and ascension > 4 mm in four patients who had no other sign of loosening. Complications were: dislocation (n = 15), phlebitis (n = 2), sciatalgia (n = 1) and infection (n = 1). Discussion This retrospective series revealed an unacceptable dislocation rate with no correlation with implant positioning error in comparison with the biomechanical position of the healthy hip. The non-hemispheric shape of the implant and the beveled edge of the polyethylene insert could be implicated, particularly since none of the implants had a polyethylene overhang. The mean annual wear (0.087 mm) was relatively high compared with cemented cups. Age was the only factor correlated with wear. Migrations were difficult to interpret. Medialization or ascension of the implant was not associated with any clinical sign suggestive of loosening. The poor reproducibility of the x-rays cannot explain these findings. Secondary repositioning of the implant might be involved. Conclusion The high rate of dislocation and the frequency of excessive wear have led us to abandon this implant for more hemispheric cups.