Subjects were tested for: free recall and judgments of self-accuracy and completeness, cued-recall and certainty of recall, visual identification, voice identification, and subjective confidence in identification performance. Three of the general guidelines articulated by the U.S. Supreme Court (Neil v. Biggers, 1972) for the determination of eyewitness accuracy were tested. Support was found for Principle I, i.e. 'the opportunity of the witness to view the criminal at the time of the crime.' However, little support was found for either Principle 3, 'the accuracy of the witness' prior description of the criminal,' or Principle 4, 'the level of certainty demonstrated by the witness at the time of confrontation' as important predictors of eyewitness accuracy. This study showed that subjects do not allocate greater attention and memory for voice identification as the quality of visual observation conditions decline. Few reliable memory differences between males and females were found, although males indicated greater certainty in cued recall. (Author abstract)