The purpose was to compare the reliability of landmark identification and linear and angular measurements in conventional versus digital cephalometry. Using 50 cephalometric radiographs, four orthodontic residents identified 19 cephalometric landmarks followed by 18 linear and angular measurements of the same radiographs. The values of 18 measurements were compared to quantify the measurement difference and interobserver errors between these two methods. Multivariate analysis of variance showed that the "cephalometric radiograph" and "landmark" variation had greater influence than that of "method" (landmark identification on original radiograph / on digital). A statistically significant difference for interobserver errors between the two methods was noted only for 5 out of 19 cephalometric landmarks. The most accurately identified landmark in conventional and digitized method was Sella (S), followed by Nasion (N). Landmarks requiring further scrutiny in digital images were Porion (P) Articulare, ANS, UM, and LM. The advantages of digital cephalometry were also substantiated. Keywords: computer-aided cephalometric analysis, digital imaging, cephalometric measurements