The results indicate that the measurements of the following anthropometric landmarks failed to identify targets accurately: the horizontal distance from the outer edge of one eye to the outer edge of the other eye. The horizontal distance from the outer edge of each eye to the center of the nose at eye level, the vertical distance between the center of the upper nose at eye level and the center of the mouth, and the angular distances from the center of the mouth to the outer edge of each eye. These findings suggest that the comparison of video images and photographs using anthropometric proportions from the chosen landmarks, even under nearly ideal conditions, is apparently of limited value in criminal identification cases. The authors suggest that future research focus on biometrics, facial structural comparisons, and general improvement in the quality of images produced from video surveillance systems. This study examined an existing database of photographic lineups, in which one video image was compared against 10 photographs. The target and test photos were of high quality, although taken with a different camera. 4 figures and 21 references