This research compared the length determination images produced by Trophy and Regam direct digital radiographic systems with conventional E-speed radiographs. Size #15 K files were placed in the canals of 19 teeth of cadaver specimens and radiographed in a geometrically standardized bench device. Three evaluators estimated the length adjustment necessary to position the file at the apex. The mean estimated adjustment lengths were compared with the true, anatomically determined adjustment lengths. Statistical analysis with analysis of variance and the Student-Newman-Keuls test indicated no significant difference between direct digital thermal print images and conventional radiographs (p > 0.10). Length adjustment estimates with conventional radiographs were significantly more accurate than Regam high-resolution computer monitor images (p < 0.01). Additionally, it was determined that length adjustment estimates were significantly more accurate when the file was placed short of the apex (p < 0.001). The clinical relevance of these differences is discussed.