The anatomical femoral insertion of the normal anterior cruciate ligament (ACL) lies on the deep portion of the lateral wall of the intercondylar fossa. Following the deep bone-cartilage border, it stretches from 11 o'clock high in the notch all the way down to its lowest border at 8 o'clock. The tension curve of the normal ACL during passive flexion-extension shows a characteristic pattern with two tension peaks, one in full extension and the other in deep flexion. Low tension in mid-flexion is measured between those two peaks. Standard ACL reconstructions use a femoral insertion at 11 o'clock, which results in non-physiological tension curves with one peak in full extension only. The hypothesis was that it is possible to reproduce the typical tension curve of the normal ACL by placing the femoral tunnel into the low part of the normal ACL footprint. In a controlled laboratory study, three different femoral tunnel positions at 9, 10 and 11 o'clock were tested in four cadaver knees while using the identical standard tibial tunnel each time. The tension curve was measured during passive flexion-extension with a custom-made tension measurement device, using an artificial Dacron graft. The tension curve of grafts in the 9 o'clock tunnels showed the characteristic pattern of the normal ACL's tension curve with tension peaks in extension as well as in deep flexion. The tunnels in the 10 and 11 o'clock positions failed to reproduce the normal ACL's tension curve with high-tension values in extension only. If an ACL graft is placed low within the normal femoral footprint, the physiological tension curve of the normal ACL can be reproduced.