Purpose: To investigate cone beam computed tomography (CBCT) accuracy in measuring facial bone height and detecting dehiscence and fenestration defects around teeth. Materials and Methods: Patients who were treatment planned for periodontal flap or dental implant surgeries were enrolled (n = 25). CBCT imaging (Carestream CS 9300) was obtained at 0.09-mm voxels (n = 10 patients, 23 teeth) and at 0.18-mm voxels (n = 15 patients, 33 teeth). Facial bone height measurements, from cusp tip to crest of bone height along the long axis of the tooth, and presence or absence of dehiscence or fenestration defects were recorded from CBCT images in triplicates independently by two examiners. The corresponding clinical measurements were made at the time of surgery. Comparisons of CBCT and clinical measurements were made using paired t tests for teeth: anterior and posterior, maxillary and mandibular, with or without restorations, or root canal therapy. Level of agreement between investigators was assessed by concordance correlation coefficients (CCC), Pearson's correlation coefficient (PCC), and Cohen's Kappa. Results: Comparing mean CBCT and clinical measurements, statistically significant differences were noted for 0.09-mm and 0.18-mm voxel sizes, for anterior and posterior teeth, for maxillary and mandibular teeth, for teeth with or without restorations, and for teeth without root canal therapy (P .05). Clinical and CBCT measurements were similar for teeth with crowns and with root canal therapy (P > .05). CBCT measurements underestimated mean facial bone height from 0.33 ± 0.78 to 0.88 ± 1.14 mm (mean ± SD) and absolute facial bone height values from 0.56 ± 0.35 to 1.08 ± 0.92 mm. Intraexaminer and interexaminer reliability for measuring facial bone height ranged from poor to substantial (PCC = 0.78 to 0.97 and CCC = 0.63 to 0.96, respectively). Interexaminer reliability for detection of dehiscence and fenestration defects ranged from poor to moderate (Cohen's Kappa = -0.09 to 0.66). Conclusion: CBCT imaging underestimated facial bone height and overestimated the presence of dehiscence and fenestration defects. Palabras clave: bone, cone beam computed tomography, diagnostic imaging, radiography