The effects of direct currents on internal remodeling were examined using femurs of 21 adult mongrel dogs. The left femurs of all dogs, used as a control, were inserted electrodes only, which were not electrically stimulated. In Group 1, no surgery was done on the right femur. In Group 2, 1 microA continuous direct current (D. C.) was applied for 6 weeks; in Group 3, 10 microA for 2 weeks; in Group 4, 10 microA for 4 weeks; in Group 5, 10 microA for 6 weeks; in Group 6, 10 microA for 16 weeks; and in Group 7 100 microA for 6 weeks. Each group was composed of three dogs. The following histomorphometric parameters were measured to evaluate the effects of electrical stimulation: number of resorption cavities (Ar), number of secondary osteons with osteoid seam (osAf), linear rate of mineralization of osteoid seam (Mo), perimeter of osteoid seam (Sf), cross-sectional area of secondary osteon (Ah) and bone formation rate (Vf). The following results were obtained: In Group 1, resorption cavities and secondary osteon with osteoid seam were observed more in the left femur than in the right. Then, mechanical stimulation of periosteal stripping or cortical drilling enhanced internal remodeling of cortical bone. In other groups (electrical stimulation was applied on the right femur), it seems that internal remodeling, especially activation frequency, was enhanced by D. C. No significant change was noted in the linear rate of mineralization of osteoid seam and cross-sectional area of the secondary osteons. Bone formation rate in the right femur showed increment when length of stimulating period had increased from 2 to 6 weeks. In Group 6 and 7, the enhanced area of bone formation has increased in cross-section, although in Group 7, bone necrosis was observed around the electrode in the right femur. Therefore, the optimum current of electrical stimulation may be just or slightly more than 10 microA. Bone formation rate was correlated with volume of callus in the marrow cavity within the period of six weeks.