Healing of a tendon graft to a bone tunnel is slower than the healing of a bone plug. Therefore, the device chosen for hamstring fixation may need to maintain its strength and stiffness longer than the device chosen for bone-tendon-bone fixation. We evaluated, in an extraarticular ovine model, how 4 and 12 weeks of implantation affect the strength of a tendon graft fixed to bone with the Evolgate. The long digital extensor tendon was transplanted and fixed with the Evolgate into a 30-mm long, 8 mm diameter bone tunnel drilled in the tibial metaphysis of both posterior limbs of 15 skeletally mature Suffolk sheep. Immediately after implantation, and 4 and 12 weeks later, biomechanical cyclic load tests in 50 N increments were performed until failure to evaluate the ultimate failure load (UFL). Histological analysis was also performed at 4 and 12 weeks. Biomechanical tests revealed a UFL of 339±120 N at time 0, and increases to 635±19 N (4 weeks) and to 867±80 N (12 weeks). The differences between all 3 groups were significant (p<0.001, paired t test). The histological evaluation showed a layer of cellular, fibrous tissue between the tendon and the bone, along the length of the bone tunnel; this layer progressively matured and reorganized during the healing process. The collagen fibers that attached the tendon to the bone resembled Sharpey's fibers. The strength of the interface significantly and progressively increased between weeks 4 and 12 after transplantation, and was associated with a degree of bone ingrowth noted histologically. The use of the Evolgate seems not to interfere with the bone ingrowth after implantation, allowing an improvement in strength of the bonetendon-device complex.