OBJECTIVE The objective of this study was to evaluate the accuracy of alveolar crestal bone detection in a comparison of unenhanced and enhanced Sidexis (Siemens Medical Systems, Inc., Bensheim, Germany) digital images with Ektaspeed Plus (Eastman Kodak, Rochester, N.Y.) films by means of receiver operating characteristic analysis. STUDY DESIGN More than 100 proximal and furcal areas in the anterior and posterior areas of the mandible and maxilla of each of three tissue-equivalent human skull phantoms were imaged with film, direct digital images, and contrast- and brightness-enhanced digital images. Alveolar crest status was assessed by a consensus panel of three experts who used the Delphi method and evaluated information from all of the imaging modalities concurrently. Five observers assessed all images for the presence or absence of crestal bone loss using a five-point confidence scale. Receiver operating characteristic curves were generated, and calculated areas (Az) were analyzed by means of analysis of variance. RESULTS The expert panel determined that of 106 crestal areas, 48 were disease free and 58 exhibited bone loss greater than 2 mm apical to the cemento-enamel junction. Analysis of variance of observer Az scores showed significant differences among readers (p < 0.001) and among readings (p = 0.027), but not among modalities (p = 0.435). Mean Az values for the different modalities were as follows: Sidexis, 0.70; enhanced Sidexis, 0.71; Ektaspeed Plus films, 0.735. CONCLUSION The Sidexis digital imaging system was not significantly different from Ektaspeed Plus film for crestal bone evaluation in this in vitro study.