The research program designed to study the extent to which memory for faces can be altered used 521 college students. Pilot results indicated that subjects who viewed a face and then heard a description of that face ostensibly written by another witness were influenced by that description. Subjects who heard a misleading detail tended to incorporate that detail into their reconstruction of the original face; subjects who did not hear the misleading detail rarely did so. Three further experiments entailed subjects viewing a target individual in a film, in a photograph, or live. Some subjects were then exposed to misleading information about the target individual ostensibly given by another witness. Over a third of the subjects viewing the film included the misleading detail in their own description, using the exact wording that the other witness used; control subjects rarely did so (5 percent). Moreover, nearly 70 percent of the subjects seeing the photograph later 'recognized' an individual with the feature mentioned by the witness; control subjects did only 13 percent of the time. In the live experiment, if subjects were questioned with leading questions containing misinformation, over 30 percent indicated they had seen the feature; only 4 percent of the control subects mentioned it. A note and 11 references are given. (Author abstract modified)