While there is no empirical evidence that a certain physical appearance is necessarily related to a particular type of criminality, there are indications that the public's stereotyping of particular physical characteristics in relation to particular criminal offenses produces a labeling phenomenon that generates a self-fulfilling prophecy for those having the negative characteristics. In testing the existence of such stereotyping, 10 photographs of male noncriminals between the ages of 27 and 33 were shown to 48 adult members of the general public and 10 police officers, and they were asked to match the photographs with listed crimes. Statistical analyses (KolmogorovSmirnov tests) of the data revealed that for the crimes of arson, theft, rape, and burglary, no face was chosen significantly more frequently than any other; however, for the crimes of mugging, violent robbery, company fraud, soliciting, illegal possession of drugs, and gross indecency, particular photographs were chosen significantly more often than the others. No effects of age or sex of the observer were significant in the matching of photographs with offenses. While the data from the police observers comes from too small a sample to be analyzed statistically in the same way as the data from the general public sample, the conclusions drawn from the police data are similar to those derived from the general public sample. Further research should examine the effects such stereotyping has on such circumstances as suspect identification, social interaction tending to produce deviancy, and the use of police discretion. Tabular data and 12 references are provided.