Orthopedic injuries often require surgical reattachment of tendon to bone. Tendon ends can be sutured to bone by direct apposition to the bone surface or by placement within a bone tunnel. Our objective was to compare early healing of a traditional surface versus a novel tunnel method for repair of the flexor digitorum profundus (FDP) tendon insertion site in a canine model. A total of 70 tendon-bone specimens were analyzed 0, 5, 10 or 21 days after injury and repair, using tensile and range of motion mechanical testing, histology and densitometry. Ultimate force (a measure of repair strength) did not differ between surface and tunnel repairs at day 0. Both repair types had reduced strength at 10 and 21 days compared to 0 days, indicative of deterioration of suture grasping strength (tendon softening). At 21 days, tendons repaired in a bone tunnel had 38% lower ultimate force compared to surface repairs (p ¼ 0.017). Histological findings were comparable between repair groups at 5 and 10 days but differed at 21 days, when we saw evidence of maturation of the tendonbone interface in the surface repairs compared to an immature fibrous interface with no evidence of tendon-bone integration in the tunnel repairs. After accounting for bone removed by the tunnel, no difference in bone mineral density or trabecular bone volume existed between surface and tunnel repairs. If the results of our animal study extend to healing of the human FDP insertion, they indicate that FDP tendons should be reattached to the distal phalanx by suture to the cortical surface rather than suture in a bone tunnel. ß