Information obtained from police reports were used to analyze the 680 robberies that occurred at these stores between 1986 and 1988. The researchers used 18 locational or security protection independent variables including, inter alia, speed limit near the store, number of lanes on the adjoining street, number of parking lot entrances, distance from main and side streets, corner strip or shopping complex, entrance and perimeter lighting, number of employees on duty, visibility through exterior glass, location of checkout counter, mirrors on aisles, entrance buzzer, and in-store camera. A multiple regression model was constructed to explain variations in the frequency of robberies at the stores. According to the findings, the model explained only 30 percent of the variance. Only 6 of the 18 variables were significantly related to the frequency of robberies and only two -- corner strip or shopping complex and number of employees on duty -- were related in the hypothesized direction. 2 tables, 34 notes, and 79 references