The purpose of this study was to determine whether bone mineral density of the host bone, measured using conventional dual photon absorptiometry techniques, and insertion torque can predict part of the ultimate failure strength of interference screw fixation of quadrupled hamstring tendon grafts. The semitendinosus and gracilis tendons were harvested from 10 human cadaveric knees, mean age 66.5 years (range, 53 to 81). The bone tunnel was sized within 0.5 mm of the graft. The graft was fixed with a biodegradable screw (7 x 25 mm for the femur, and 9 x 25 mm for the tibia) directly against the tendon and at the joint surfaces. Tibial fixation and femoral fixation were tested to failure using a materials testing system. Bone mineral density was measured in the metaphyseal region of the tibia and femur. The results of multiple regression analyses showed that both insertion torque and bone mineral density were related to the maximum load the graft withstood. These two variables explained 77.1% of the maximum load observed. We concluded that bone mineral density measurements of the host bone site are an important determinant of postoperative graft strength and thus have an important, but previously unrecognized, clinical role in establishing individual postsurgery rehabilitation protocols. Insertion torque in this study was a useful predictor of graft fixation strength.