This manuscript describes Active Appearance model for automatic Cephalometric analysis in forensic science, which is concerned with the recognition, identification, individualization, and evaluation of physical evidence. A complete model of shape and texture was built from a dataset of manually annotated images, and then tested with unseen images. To apply AAM Method for identifying landmark points 140 images were collected from AAOF out of it 67 were female and 73 were male with age ranging from 15 years to 45 years. By using the AAM algorithm, the mean shape is extracted and the appearance variation collected by establishing a piece-wise affine warp between each image of the training set and the mean shape