The purpose of this study was three-fold: to reexamine whether periapical and central lesions can be detected on the radiograph when they are in the cancellous bone, to investigate regianal differences in radiograph visualization of the lesion, and to. compare the interpretation of radiagraphs by densitometric analysis with the visual interpretation af the same radiographs. Forty-nine apical lesions and eighty-two central lesions were made in 8 mandibles and radiographed serially while being enlarged and deepened. All radiagraphs were evaluated by three dental radiologists. Each lesion was scored of 3 to 0, based on the radiographic appearances. Special radiographs of 15 apical and central lesions were analysed by densitometer and the results were compared with the scores which were previously assigned by three examiners. From this study the following conclusians could be drawn; 1. Experimental lesions confined to cancellous bone were detected an the radiagraphs in 41.3% of the cases. 2. Visualization of periapical lesions in cancellous bone was more apparent than that of central lesions in cancellous bone. 3. Visualization of the periapical lesions confined to. cancellous bane was highest in molar teeth, and detection of the central lesions in alveolar portion was easier than those in lower portion of mandibular body area. 4. It was possible to detect the experimental lesions without cortical plate involvement in 78.6% of the all lesions. 5. Densitometric analysis of radiographs could reproducibly distinguish bone changes, especially of central lesions in cancellous bone.