Background and Objectives: Detection of fenestration and dehiscence is clinically difficult and requires radiography. This study aimed to assess the effect of voxel size and resolution on the detection of fenestration and dehiscence on cone-beam computed tomography (CBCT) scans. Materials and Methods: In this diagnostic accuracy study, eight fenestration and dehiscence defects were artificially created in the buccal bone plate of incisors, premolars and molars in a dry human mandible with teeth. The defects were created randomly. This model served as the gold standard. The holes were covered with two layers of boxing wax for soft tissue simulation. CBCT scans were taken with 0.4, 0.2 and 0.15 mm resolution. The obtained CBCT scans were mixed and randomly observed by two experienced oral and maxillofacial radiologists. The observers were requested to identify the teeth with defects and specify the type of defect as well (no defect, dehiscence or fenestration). Data were analyzed using SPSS. Results: The sensitivity for detection of fenestration was 0.37-1 at 0.4 mm, 0.10-1 at 0.2 mm and 0.45-1 at 0.15 mm resolution. The specificity for detection of fenestration was 0.96-1 at 0.4 mm, 0.36-1 at 0.2 mm and 0.73-1 at 0.15 mm resolution. The specificity for detection of dehiscence was 0.95-1 at 0.4 mm and 0.15 mm resolutions and one at 0.2 mm resolution. The inter-rater agreement was 0.9-1 for all three resolutions, while the intra-rater agreement was 0.9-1. Conclusion: The use of smaller voxel size to increase resolution is not recommended to detect fenestration and dehiscence.