The following hypothesis was tested: the location of the tibial and femoral anterior cruciate ligament (ACL) attachments will differ according to the measurement technique (plain radiographs or CT-scan) in relation to the anatomic frame of reference. 10 gross specimens were studied. The location of metallic reference pins implanted around the tibial and femoral ACL attachments was recorded with reference to the bone contours with a caliper on the anatomic preparation, with standard plain AP and lateral radiographs and with a CT-scan. Results were compared with appropriate statistical tests at a 0.05 level of significance. The mean ratio between the antero-posterior tibia measurement and the center of the ACL tibial attachment varied from 50 to 52%. The mean ratio between the medio-lateral tibia measurement and the center of the ACL tibial attachment varied from 49 to 51%. The mean ratio between the antero-posterior femur measurement and the center of the ACL femoral attachment varied from 74 to 80%. The results were significantly different between the three techniques (respectively P=0.003, P=0.02 and P=0.045). The paired differences were small (3% at the tibia, 6% at the femur). There was a strong correlation and a good agreement between the three techniques. It is likely that the small differences on the tibia have few, if any, clinical relevance. Both radiographic and CT-scan measurement techniques used during the present study have the potential to be used as quality control after ACL replacement. Diagnostic study-investigating a diagnostic test. Development of diagnostic criteria in a consecutive series of patients and a universally applied "gold" standard, Level II.