The specificity of memories has been identified as a factor affecting reality monitoring performance. To examine the reality monitoring model of Johnson and Raye (1981) and to explore the relationship between memory specificity and reality monitoring, the amount of cognitive operations involved in processing information was manipulated for 72 subjects who imagined and perceived spatial relationships between objects in treasure maps. Subjects (N=24) in the whole map condition saw a continuous path between five objects, while subjects (N=48) in the segmented map condition saw segments of the path which they mentally combined into a continuous path. All subjects imagined spatial relationships by reading descriptions of imaginery hands of a clock. All subjects perceived ten maps and imagined ten maps. After presentation of the maps, the accuracy of subjects' memory for the relative location of objects was tested and subjects completed a forced-choice recognition and identification of origin task. An identification of origin score computed for each subject reflected the accuracy with which subjects discriminated between memories of imagined and perceived maps. The results indicated that subjects in the whole map condition were significantly more accurate at correctly identifying the origin of maps than were subjects in the segmented map condition. The findings demonstrated that difference in the amount of cognitive operations was a reliable reality monitoring cue and differences in the cognitive operations produced the opportunity for differences in memory specificity to be a reliable reality monitoring cue. (NRB)