This study evaluated extreme exposure reduction in cephalography with monitor-displayed DICOM (Digital Imaging and Communication in Medicine, Rosslyn, Va) images. The purposes were to investigate to what extent exposure can be reduced without losing diagnostic information for tracing analysis, whether postprocessing noise reduction could benefit localization of anatomic landmarks in extreme low-exposure images, and the quality of the images. Three images (optimal, extreme low exposure, and noise-reduced extreme low exposure) were obtained from 24 volunteers. Three orthodontists traced 18 landmarks and subjectively judged the quality of the images. Sixteen of the landmarks were well identified on the extreme low-exposure images when they were compared with the optimal images. Apex inferius and apex superius were not well identified. The evaluation of image quality on a visual analogue scale showed that the optimal images were given the highest scores, extreme low-exposure images with postprocessing noise reduction the second highest scores, and extreme low-exposure images the lowest scores. Monitor-based tracing analysis on storage phosphor cephalographic images with a 94% exposure reduction, compared with conventional exposure, can be performed with only 2 of 18 landmarks not well identified. The landmarks were better identified on the low-exposure images with standard postprocessing than on the images with post-processing noise reduction, although the subjective evaluation of the image quality indicated the opposite.