In recent years the problem of misidentification, especially in criminal cases, has attracted a good deal of attention from the media, lawyers, psychologists, and others. One influential, but by no means universally held, view of the problem might be stated in some such terms as these: from time to time an innocent man is convicted by a jury of a crime he did not – or probably did not – commit, on the basis, mainly or entirely, of eyewitness testimony relating to identification. Typically this testimony is honest, but mistaken. It may be only half-a-dozen or so chaps a year who suffer such miscarriages of justice, and some of these are professional criminals who might well have been put inside for some other offence; no system of criminal justice can be expected to eliminate all mistakes, but our legal tradition and public opinion place a high value on safeguarding the innocent, even at the price of letting some, but not an unlimited number, of criminals go free. The problem peculiar to identification is that the value of the evidence is exceptionally difficult to assess. The task is to reduce the risk of error by taking measures to improve the law and procedure governing evidence of identification. In this view the main relevance of psychology to the problem of misidentification is thought to relate to the cognitive processes of individual witnesses (notably perception, attention, memory, bias, and suggestion) rather than to cognitive processes of other participants and interactive aspects of legal processes.