Purpose: The aim of this study was to evaluate the accuracy of implants inserted using a mucosa- supported stereolithographic surgical guide and to determine the influence of surgical management of the guide (fixed or unfixed), arch (maxilla or mandible), and smoking habit (normal or hyperplastic mucosa) on accuracy. Materials and Methods: In completely edentulous subjects, preoperative computed tomography (CT) was performed, and the images were used to plan implant positions. After the implants were placed, CT was performed again, and the presurgical and postoperative images were compared. With computer software, the jaw contours from the two CT scans were matched and the deviations between the planned and actual implant positions were evaluated. Surgical technique, arch, and smoking habit were examined as independent variables, and their influence on accuracy was evaluated with a t test. Results: Twenty-eight surgical guides (225 implants) were included in this study. Deviations between planned and actual positions were seen in the global coronal (mean ± SD: 1.68 ± 0.6 mm), global apical (2.19 ± 0.83 mm), and angular (4.67 ± 2.68 degrees) dimensions. Fixation of surgical guides (fixed: 4.09 degrees; not fixed: 5.62 degrees) and use of the guide in the maxilla (4.36 degrees; mandible: 5.46 degrees) resulted in statistically significantly less angular deviation (ie, better accuracy). Nonsmoking patients showed statistically significantly better accuracy in global coronal (nonsmokers: 1.54 mm; smokers: 1.83 mm) and global apical (nonsmokers: 2.08 mm; smokers: 2.27 mm) deviations. Conclusion: The greater supporting surface of the maxilla and fixation of the surgical guide improved the accuracy of the guides. The reduced mucosa thickness in nonsmokers decreased global coronal and global apical deviation. Keywords: accuracy, guide, computer-assisted image processing, dental implants, edentulous, stereolithography, surgical guide, three-dimensional imaging