In the present study, we explore how reading habits (e.g., reading from left to right in French or reading from right to left in Arabic) influence thescanningand theconstructionof mental models in spatial reasoning. For instance, when participants are given a problem like A is to the left of B; B is to the left of C, what is the relation between A and C? They are assumed to construct the model: A B C. If reading habits influence the scanning process, (...) then readers of French should inspect models from left to right, whereas readers of Arabic should inspect them from right to left. The prediction following this analysis is that readers of French should be more inclined to produce “left” conclusions (i.e., A is to the left of C), whereas readers of Arabic should be more inclined to produce “right” conclusions (i.e., C is to the right of A). Furthermore, one may expect that readers of French show a greater ease in constructing models following a left-to-right direction than models following a right-to-left direction, whereas an opposite pattern might be expected for readers of Arabic. We tested these predictions in two experiments involving French and Yemeni participants. Experiment 1 investigated the formulation of conclusions from spatial premises, and Experiment 2, which was based on non-linguistic stimuli, examined the time required to construct mental models from left to right and from right to left. Our results show clear differences between the two groups. As expected, the French sample showed a strong left-to-right bias, but the Yemeni sample did not show the reverse bias. Results are discussed in terms of cultural influences and universal mechanisms. (shrink)