The first study involved the development of scripts for the typical actions involved in the robbery of a convenience store. There was much agreement among the 120 college student participants concerning the sequence of actions for such a robbery. The second study was based on the information obtained in the first study and aimed to determine if the script would generate gap-filling errors in memory similar to those obtained in other script studies. The 144 participants were divided into 8 groups. Each participant viewed one of two sequences of color slides depicting a robbery of a convenience store by a lone robber. One sequence omitted three central script actions; the other sequence omitted three peripheral script actions. In addition, the rate of exposure was varied (2 versus 8 seconds), as was the length of the retention interval (5 minutes versus 1 week). Results confirmed predictions of a higher rate of false recognition for central actions than for peripheral actions; this tendency was exaggerated for the longer retention interval. Findings indicated that people have expectations of what happens in an event such as a robbery, and that scripts can be a source of gap-filling errors in eyewitness accounts. Further research should address the limitations of this study and explore additional factors that influence the memory of eyewitnesses. Table, figure, and 24 references (Author abstract modified)