The aim of this study was to examine the effect of long-term locking plate fixation on the cortical bone of the canine radius. Locking compression plates were fixed to the left and right radius in dogs (n = 3). The left radius was fixed with a locking head screw (Locking Plate group, LP). The locking compression plate was compressed periosteally in the right radius using a cortex screw (Compression Plate group, CP). Radial bones from dogs that were euthanized for other purposes were collected as an untreated control group (Control group). After euthanasia at 36 weeks following plate fixation, radial bones were evaluated for bone mineral density and underwent histological analysis. Bone metabolic markers were analyzed by quantitative polymerase chain reaction (qPCR). Statistical analyses were performed for comparisons between groups. The LP group showed no significant difference in bone mineral density after plate fixation, whereas the CP group showed significantly lower bone mineral density. Histological analysis indicated that the number of osteoclasts and rate of empty lacunae increased significantly in the CP group relative to the Control and LP groups. qPCR analysis indicated increased expression of inflammatory cytokines, such as tumor necrosis factor-alpha, interleukin-6, and tumor necrosis factor ligand superfamily member 11 in the CP group, whereas Runt-related transcription factor 2, an osteoblast marker, was similar in all groups. The expression of hypoxia-inducible factor-1[alpha] in the CP group was also increased relative to that in the Control and LP groups. Thus, locking plate fixation is a biologically superior fixation method that does not cause implant-induced osteoporosis in the bone in the long term.