The aim of this study was to evaluate and compare the repair of bone defects filled with calcium aluminate cement (EndoBinder), mineral trioxide aggregate (MTA), and calcium hydroxide. Methods After mixing, the cements were inserted into bone defects (3.3 mm) mechanically created in the right and left tibias of 30 rats (Rattus norvegicus, Wistar). In the control group, the bone defects were filled with blood clot of the animal itself. After time intervals of 7, 30, and 90 days had elapsed, bone tissue biopsies (n = 5) were surgically obtained and submitted to laboratory processing. The response of bone tissue in contact with the materials was microscopically analyzed. The percentage of neoformed bone tissue in the defect was determined by means of planimetry counting points superimposed on the histologic image. Results Significant increase in the percentage of neoformed bone tissue was observed throughout the experimental periods in all groups (P < .05). For the cements EndoBinder and MTA (30 and 90 days), these percentage values were statistically higher than those of the control group (P < .05); however, they were similar to those of calcium hydroxide (P > .05). Conclusions EndoBinder and MTA allowed complete repair of bone defects created in rat tibias.