The current study investigated how computer games and maps compare as preparation for readers to comprehend and retain spatial relations in text narratives. Readers create situation models of five dimensions: spatial, temporal, causal, goal, and protagonist (Zwaan, Langston, & Graesser 1995). Of these five, readers mentally model the spatial dimension least well (Rinck, 2005). Studying maps before reading improves retention of general details from non-narrative readings (Kulhavy & Stock, 1996). The current study investigated how playing interactive computer games compared with studying a computer-based map as a preparation for reading a narrative. The dependent variables were: 1) evidence of monitoring spatial relations while reading stories, and 2) comprehension and retention of spatial relations in stories. Evidence of monitoring of the spatial relations was measured by average times, in milliseconds, for reading individual sentences with changes in protagonist location. Comprehension and retention of spatial relations in stories were measured by multiple-choice posttests of spatial relations in the stories. Eighty 11-year-olds participated in all three experimental conditions: 1) studying a map with sound and animations but no interaction, 2) playing an interactive computer game, and 3) completing a filler task. Each condition was followed by reading a narrative and then taking a spatial posttest. In terms of multiple-choice posttests, map condition had the highest average number correct, closely followed by the computer game. Filler task condition was a distant third. No between-condition differences were found for the reading times on sentences with changes in protagonist location. Results suggest that maps may be superior to computers games as preparation for spatial reading. (Contains 4 tables and 3 figures.)