Purpose: The different implant systems available today present several types of surface treatment, with the aim of optimization of bone-implant contact. This study compared 4 different types of implant surfaces. Materials and Methods: The first, second, third, and fourth mandibular premolars were extracted from 5 young adult mongrel male dogs. Ninety days after removal, four 3.75-mm-diameter, 10-mm-long screw-type implants (Paragon) were placed with different surface treatments in mandibular hemiarches. The dogs received 2 implants of each of the following surface treatments: smooth (machined), titanium plasma spray (TPS), hydroxyapatite coating (HA), and sandblasting with soluble particles (SBM). The implants were maintained unloaded for 90 days. After this period, the animals were sacrificed, and the hemimandibles were extracted and histologically processed to obtain nondecalcified sections. Two longitudinal ground sections were made for each implant and analyzed under light microscopy coupled to a computerized system for histomorphometry. Results: The following means were obtained for bone-implant contact percentage: machined = 41.7%, TPS = 48.9%, HA = 57.9%, and SBM = 68.5%. Discussion: The means for all treatments that added roughness to the implant surface were numerically superior to the mean found for the machined surface. However, this difference was statistically significant only between groups SBM and machined (Tukey test, P .05). Conclusions: The SBM-treated surface provided a greater bone-implant contact than a machined surface after 90 days without loading in this model.