Purpose: Recent clinical studies have shown that implant placement is highly predictable with computergenerated surgical guides; however, the reliability of these guides has not been compared to that of conventional guides clinically. This study aimed to compare the accuracy of reproducing planned implant positions with computer-generated and conventional surgical guides using a split-mouth design. Materials and Methods: Ten patients received two implants each in symmetric locations. All implants were planned virtually using a software program and information from cone beam computed tomographic scans taken with scan appliances in place. Patients were randomly selected for computer-aided design/computer-assisted manufacture (CAD/CAM)-guided implant placement on their right or left side. Conventional guides were used on the contralateral side. Patients underwent operative cone beam computed tomography postoperatively. Planned and actual implant positions were compared using three-dimensional analyses capable of measuring volume overlap as well as differences in angles and coronal and apical positions. Results were compared using a mixed-model repeated-measures analysis of variance and were further analyzed using a Bartlett test for unequal variance (α = .05). Results: Implants placed with CAD/CAM guides were closer to the planned positions in all eight categories examined. However, statistically significant differences were shown only for coronal horizontal distances. It was also shown that CAD/CAM guides had less variability than conventional guides, which was statistically significant for apical distance. Conclusion: Implants placed using CAD/CAM surgical guides provided greater accuracy in a lateral direction than conventional guides. In addition, CAD/CAM guides were more consistent in their deviation from the planned locations than conventional guides.