The purpose of this study was to assess reproducibility of cephalometric measurements in cephalograms obtained by three dentomaxillofacial radiology clinics. Forty lateral cephalometric radiographs were selected and sent at different times to three different clinics for cephalometric analyses. Each clinic digitized the radiographs with the same resolution, and landmarks were located with the mouse pointer directly on the digitized radiographic image on the screen. Three cephalograms were obtained from each radiograph, totaling 120 analyses. Data were analyzed with analysis of variance. Of the 32 factors studied, reproducibility of results was satisfactory for only four factors: position of maxilla relative to anterior cranial base, inclination of occlusal plane relative to anterior cranial base, position of lower incisor relative to nasion-pogonion line, and soft-tissue profile of face (P < .05). Differences in cephalometric measurements were present and such differences were significant for most factors analyzed. The different cephalometric measurements obtained by the three dental radiology clinics were not reproducible.