Between 1986 and 1993, as part of a prospective on-going trial, 205 patients had prosthetic ACL reconstruction for chronic symptomatic knee instability. The mean interval from injury to surgery was 39.2 (range 6-216) months.Overall, there were 22 (10.7%) confirmed ruptures and 183 satisfactory implants. The mean follow-up was 30.4 months in the ruptured group and 30.2 months in the satisfactory group.Between the two groups, there was no significant difference in the cause of the injury, the side involved, the sex of the patient, associated intra-articular pathology, pre-injury Tegner score and preoperative Lysholm or Tegner scores.The mean age of the 22 patients in whom the ligament ruptured was 24.5 years, significantly younger (P > 0.001) than the satisfactory group in which the mean age was 30.0 years. Of the 22 patients who ruptured, 20 (91%) were under 30. In the 183 satisfactory implants, 101 (55%) were under 30. This association between age and implant failure is significant (P < 0.01).In both groups, Lysholm and Tegner scores significantly (P < 0.001) improved over preoperative scores. However, the 183 patients with a satisfactory ligament obtained a mean Lysholm score of 89.1, significantly (P < 0.001) better than the mean score of 74.8 in the failed group. There was no significant difference in the postoperative Tegner scores of the two groups.We believe that the main reason a prosthetic cruciate ligament will fail is because of impingement. Other factors, such as the age of the patient or pre-injury activity levels also need to be taken into account.We conclude that prosthetic rupture is more likely in the younger patient and rupture resulted in significantly lower Lysholm scores after surgery. Implantation of a prosthetic ACL should perhaps be limited to patients over the age of 30 years.DiscussionQuestion. How was rupture diagnosed?Answer. Clinically initially, then by arthroscopy or more recently by MRI. KT-2000 and other laxity measurements were also made. Measurements were however not indicative of symptoms. Lysholm and Tegner score were not good at predicting rupture.Question. If the tibial placement is so far back that the ligament is vertical how can it affect instability in the horizontal plane?Answer. All prosthesis are straight, one must put it in the back of the footprint as far as possible and compromise stability in order to avoid impingement.Question. Can we diagnose partial rupture of the prosthesis?Answer. It is very difficult, markers break up themselves and do not give a good evidence of breakage. Looking at the MRI may be helpful but can be difficult.