Purpose: The aim of this study was to investigate the accuracy of linear measurements from cone beam computed tomography (CBCT) images and digital panoramic radiographs at various implant sites. Materials and Methods: Fifty implant sites from six skulls were marked with gutta-percha and subjected to CBCT with five different voxel protocols: 0.125 mm, 0.160 mm, and 0.250 mm with the 3D Accuitomo 170 CBCT machine and 0.200 mm and 0.300 mm with the CS 9500 CBCT machine. Images were also taken with the CS 9000 panoramic machine with three protocols: normal head, chin-up, and chin-down positions. Electronic linear measurement of bone height using the corresponding machine's software was recorded by two observers. Physical measurement using a digital caliper with ± 0.02-mm accuracy was directly recorded at the corresponding regions as the gold standard. All image measurements were compared with the physical measurements. The paired sample correlations for physical measurement, mean difference, standard deviation, absolute error, absolute percentage error, and inter- and intraobserver reliability were calculated. Results: Intraobserver and interobserver reliability was more than 0.99. Paired sample correlation between all image measurements and physical measurements was considered statistically significant at P .05. All image measurements were underestimated by less than 2 mm, except for the chin-down position of the maxilla in the panoramic radiograph. The absolute error and absolute percentage error in the mandible were less than those in the maxilla, and values obtained with CBCT were less than those from panoramic radiographs. Conclusion: CBCT images using the 3D Accuitomo 170 and CS 9500 machines and digital panoramic radiographs via a picture archiving and communication system are sufficiently accurate for vertical linear measurements in dental implant treatment planning. Keywords: CBCT, cone beam CT, dental implant, digital panoramic radiography, linear measurement accuracy, PACS