PURPOSE: Modifications in implant surface topography and chemistry may alter the early bone response at different levels. This study characterized four implant surfaces and evaluated the biomechanical fixation and histologic response at early implantation times in a canine radius model. MATERIALS AND METHODS: External-hexagon Branemark-type implants were used with four experimental surfaces: microblasted (MI), acid-etched and microblasted (AAM), anodized (A), and discrete crystalline deposition (DCD). Surface topography was assessed by scanning electron microscopy, interferometry, and x-ray photoelectron spectroscopy. The implants were placed in the central region of the radii of eight beagle dogs and remained in vivo for 10 or 30 days. The implants were torqued to interface failure, and a general linear statistical model with torque as the dependent variable and implant surface and time in vivo as independent variables was used. RESULTS: All surfaces presented were textured, and different surface chemistries were observed. No significant differences between implant surfaces were observed for torque at 10 days. However, at 30 days, the AAM surface presented significantly higher torque values compared to the DCD and A surfaces. Significantly higher torque values were observed at 30 days compared to 10 days (P < .001). CONCLUSIONS: Significantly different biomechanical fixation dependent on surface preparation was observed after 30 days, and all surfaces were biocompatible and osteoconductive.