Compared with conventional films digital radiography allows a wide range of exposure and scanner settings. The aim of this study was to investigate the diagnostic accuracy of the evaluation of tooth fractures depending on variations of exposure dose, scanner settings and surrounding tissues. Extracted human teeth were exposed separately and in a pig bone phantom before and after artificial fracture at 70 kV (Gendex Oralix DC) and three exposure settings using the Vistascan System I at three different resolution settings. Five dental observers evaluated 432 radiographs of 12 adequately fractured teeth under randomized conditions for the existence of a tooth fracture on a five-point scale. The highest value of true ratings (76.7%) with pig bone vicinity was achieved at 56 micro Gy (160 ms) and 20 lp mm(-1), followed by 71.7% at 27 micro Gy (80 ms) and 20 lp mm(-1). The worst diagnostic accuracy of 56.7% true ratings resulted from 6 micro Gy (20 ms) and 10 lp mm(-1). With radiographs of teeth in empty surroundings the rate of true decisions was in some cases significantly higher. For the diagnostics of tooth fractures with the Vistascan System I the diagnostic value seems to depend on, besides radiographic projection, the scanner resolution setting, the surrounding tissues and the exposure dose.