Abstract: Craniometry androentgenographic cephalometry were carried on three skulls to determine55 measurement points and do 53 items of measurements respectively in this paper. The over-lapping comparison of the measurement points shows that ' most (83.6%) of the points in cra-niometry can be determined on cephalometric roentgenograms precisely. Two-way analysis ofvariance indicates that there are no significant differences between the measurement results ob-tained by the two methods (P > 0.05). Because the definitions of some measurement points aredifferent in the two methods, the authors propose some measurementpoints should be addedand some experienced methods are proposed to determine some points which are difficult [olocate because of the overlapping of images in roentgenograms. Thepresent study discoveredthat according to the positions and distances of points on skull, the amplification rates can bedivided into several levels, which can basically reflect actual amplification errors. So the au-thors proposed that under the condition of standard amplification rate of 11.1%, the rates can bedivided into 11.1%, 8.0% and 4.0% to rectify the linear measurement data ofroentgenogra-phic cephalometry. The authors believe that roentgenographic cephalometry can be used in theresearch of physical anthropology and has good prospect in the future.

Key words: Craniometry; Roentgenographic cephalometry; Physical anthropology; Stomatology