In general, people are thought to be reasonably good at recognizing faces. However, it is difficult to match images of unfamiliar faces: the process of facial comparison. Facial comparison is a wide‐ranging forensic technique: it involves disciplines such as image analysis, anatomy, anthropology, photogrammetry, 3D technology, statistics, and cognitive psychology (human as a measuring instrument). Many steps in the comparison process are still mediocre, poorly or not supported by scientific data. The issues “quantification of image quality,” “human as a measuring instrument,” and “individualizing value of features” will be discussed in the context of man and machine's functioning as a measuring instrument. Research on humans as a measuring instrument in facial comparison is limited; most of the research concerns “recognition” or “rapid” comparison. Performance of current automated biometric systems is still too low to be of use for forensic purposes, although biometric systems may be useful for automated investigation of standardized image databases. The evidential value of surveillance material is severely limited by the quality of most surveillance material.