Lengthening of an anterior cruciate ligament (ACL) graft construct can occur as a result of lengthening at the sites of tibial and/or femoral fixation and manifests as an increase in anterior laxity. Although lengthening at the site of fixation has been measured for a variety of fixation devices, it is difficult to place these results in a clinical context because the mathematical relationship between lengthening of an ACL graft construct and anterior laxity is unknown. The purpose of our study was to determine empirically this relationship. Ten cadaveric knees were reconstructed with a double-looped tendon graft. With the knee in 25° of flexion, the position of the proximal end of the graft inside the femoral tunnel was adjusted by moving the femoral fixation device until the anterior laxity at an applied anterior force of matched that of the intact knee. In random order, the graft construct was lengthened 1, 2, 3, 4, and by moving the femoral fixation device distally along the femoral tunnel and anterior laxity was measured. The increase in the length of the graft construct was related to the increase in anterior laxity by a simple linear regression model. Lengthening the graft construct from 1 to caused an equal increase in anterior laxity (slope=, , ). Because an anterior laxity increase of or greater in a reconstructed knee is considered unstable clinically and because many fixation devices in widespread use clinically allow or greater of lengthening in in vitro tests, our empirical relationship indicates that lengthening at the site of fixation probably is an important cause of knee instability following ACL reconstructive surgery. Our empirical relation also indicates that an important criterion in the design of future fixation devices is that lengthening at the sites of fixation in in vitro tests should be limited to less than .