The most important implication of the results is that reducing the anxiety associated with making an identification sigificantly improves accuracy. The most consistently successful method of reducing anxiety with both adults and juveniles proved to be the use of color slides in lieu of a lineup. Also, distancing the witness from the lineup with a one-way vision screen reduced anxiety and enhanced identification performance, particularly with adult female witnesses. Results from two of the experiments involving adults indicated that women were generally more accurate than men. In experiments with children results concerning the relationship between the nature of the incident and identification performance were inconclusive, as were results concerning the effect of prior description upon identification accuracy. The demonstrated unreliability of identification evidence suggest that such evidence should always be corroborated in court by evidence of another kind. Graphic and tabular data, notes, and 19 references are provided. For related documents, see NCJ70738-47. (Author abstract modified)