This in vitro study compared the accuracy of file length measurements made on calibrated and uncalibrated direct digital images. Endodontic files of known lengths and ISO sizes were used in 10 single-rooted, relatively straight teeth within cadaver specimens. The crowns of the teeth were ground flat and an orthodontic wire of known length was secured to the coronal surface. This wire was placed mesiodistally and perpendicular to the root and served as the reference point for the file measurement and as a calibration reference length. A #20 file was hand-measured to a length that reached the apical third of each tooth. It was inserted and a radiographic image was secured. The instrument was remeasured three additional times at different lengths on the same tooth and reinserted before each image acquisition. Thus 40 digital images were acquired using a GE X-ray unit and a Schick Computed Dental Radiography (CDR) #2 sensor. These images were placed in random order, and an independent, blinded investigator determined the file lengths using on-screen calibrated and uncalibrated measurement of the CDR image with a straight-line and multiple-line measuring technique. The experimental measurements were compared with each other and with the known clinical measurements. A two-way analysis of variance indicated that there was a statistically significant difference showing that the calibrated measurements were more accurate than the uncalibrated measurements (p = 0.0001), and there was no significant difference between the straight-line and multiple-line measuring techniques (p = 0.14).