OBJECTIVE: The objective of this study was to evaluate the effectiveness of oblique radiographic projection using the intraoral paralleling technique in detecting various peri-implant bone defects. METHODS: Artificial mandibular models with appropriate radiopacity were created. An alveolar bone model without bone defects and models with 12 types of peri-implant bone defects (buccal, circumferential, and mixed types with different widths and depths) were created. A total of 273 images were obtained with orthoradial projections and 10-, 20-, and 30-degree oblique projections using a modified receptor holder. Two observers independently evaluated the images to detect bone defects. The grayscale values (GVs) of the peri-implant region and the adjacent area were measured and compared. The relationship between the GV and the observers' results was examined. The area under the curve (AUC) and inter-observer agreement were calculated. RESULTS: Circumferential and mixed bone defects were detected on the orthoradial projections, while buccal defects were not detected. However, the detection of buccal defects was markedly improved using the oblique projections. In particular, the highest detection rates were obtained using the 20-degree oblique projection. There were no significant correlations between the GV and the bone defect detection rate. The AUCs for the two observers were 0.712 and 0.669. The inter-observer agreement was 0.502. CONCLUSIONS: Compared with orthoradial projections, the use of oblique projection images greatly improved the ability of observers to detect peri-implant bone defects on the buccal side. The results provide new evidence for the selection of radiographic images in the follow-up of implant treatment.