Forty-four adult male mongrel dogs were used to compare segmental cortical freeze-dried allogeneic bone grafts with fresh autogenous, freeze-dried autogenous, and fresh allogeneic segmental cortical grafts. Group I consisted of bilateral fresh autografts as external controls; Group II, a fresh autograft on one side for internal control and a freeze-dried autogenous graft on the other side to evaluate the effect of freeze-drying on repair; Group III, a fresh autograft on one side and a fresh allograft on the other side to measure the differences between autogenous and allogeneic grafts; and Group IV, a fresh autograft on one side and a freeze-dried allogeneic graft on the other to see if freeze-drying altered the repair of allogeneic grafts. The grafts were analyzed qualitatively over a six-month period by the use of interval roentgenograms to determine the times of graft-host union and the incidence of fatigue fractures. Six months after operation, the repair processes in the four groups were compared quantitatively with respect to biological repair and physical strength using torsional stress-testing, tetracycline labeling, and microradiography. The results indicate both qualitatively and quantitatively that: (1) fresh bilateral segmental cortical autografts show reproducible characteristics, so that the canine fibula furnishes a satisfactory model (Group I); (2) freeze-drying does not inhibit the repair process per se (Group II); (3) fresh allografts are rejected in varying degrees of intensity (Group III); and (4) freeze-drying does not protect cortical allogeneic grafts from such rejection (Group IV).