The participants were 412 undergraduate students. They viewed a slide show of a picnic scene in which a stranger came and stole a camera placed on the picnic table by one of the actors. The participants came back immediately, 2-3 days later, or 1 week later and tried to identify the offender in lineups or showups with the offender present or absent. They were also interviewed about their memory of the crime scene and rated their confidence in their identification of the suspect. Across all time delays, the participants maintained a higher identification accuracy with the showup presentation. The miss rate in lineups increased from the immediate lineup to the 1-week delay, but the miss rate decreased in the 2-3-day delay. In addition, the false positive rate increased steadily across the delay conditions, especially when the perpetrator was absent. The rate of positive identifications remained relatively unchanged across the three delay conditions for showups. However, the miss rate and the correct rejection rate dropped significantly across delays. The false positive rate increased across delays, but it never rose about 13 percent. Further research should examine the impact of a 2-week delay. Tables and 35 references (Author abstract modified)