This study evaluated whether variations in education level and experience among dental staff influence the diagnostic accuracy of carious lesions on digital radiographs. Three student groups and a fourth group of general practitioners (Dentists) with more than five years of clinical experience participated in this study.The student groups were (i) dental students in their final (tenth) semester (DS-10), (ii) dental students in the sixth semester (DS-6) who just finished dental radiology training, and (iii) dental hygiene students (DHS) in their final (fourth) semester. Seven observers from each group participated. Standard radiographs of 100 extracted teeth (premolars and molars) were taken. The 28 observers evaluated the images for approximal carious lesions on a standard monitor. All evaluations were made in ambient light below 50 lux. Receiver operating characteristic curves were plotted to assess results. The standard criterion for healthy or carious lesions was a histological examination of sliced teeth. Kappa statistics evaluated intra-observer agreement. For carious lesions that had extended into the dentine, significant differences were found between (i) Dentists and all other groups, (ii) Dentists and DS-10 (p < 0.01), and (iii) Dentists and DS-6 and DHS (p < 0.001). Differences between DS-10 and DHS (p < or = 0.05) were also significant. In this study, education level and experience clearly influenced the diagnostic accuracy of approximal carious lesions that had extended into the dentine on digital radiographs.