Placement of dental implants requires precise planning that accounts for anatomic limitations and restorative goals. Diagnosis can be made with the assistance of computerized tomographic (CT) scanning, but transfer of planning to the surgical field is limited. Recently, novel CAD/CAM techniques such as stereolithographic rapid prototyping have been developed to build surgical guides in an attempt to improve precision of implant placement. However, comparison of these advanced techniques to traditional surgical guides has not been performed. The goal of this study was to compare the accuracy of a conventional surgical guide to that of a stereolithographic surgical guide. CT scanning of epoxy edentulous mandibles was performed using a cone beam CT scanner with high isotropic spatial resolution, while planning for 5 implants on each side of the jaw was performed using a commercially available software package. Five surgeons performed osteotomies on a jaw identical to the initial model; on the right side a conventional surgical guide (control side) was used, and on the left side a stereolithographic guide was used (test side). Each jaw was then CT scanned, and a registration method was applied to match it to the initial planning. Measurements included distances between planned implants and actual osteotomies. The average distance between the planned implant and the actual osteotomy was 1.5 mm at the entrance and 2.1 mm at the apex when the control guide was used. The same measurements were significantly reduced to 0.9 mm and 1.0 mm when the test guide was used. Variations were also reduced with the test guide, within surgeons and between surgeons. Surgical guidance for implant placement relieves the clinician from multiple perioperative decisions. Precise implant placement is under investigation using sophisticated guidance methods, including CAD/CAM templates. Within the limits of this study, implant placement was improved by using a stereolithographic surgical guide.