In a series of studies three solution strategies for spatial problems were identified: a mental rotation strategy, a perspective change strategy, and an analytic strategy. The test-retest reliability, validity, and general inability of a measure of these strategics was established. This measure, the Spatial Strategy Questionnaire, was then used to assess the role of solution strategy in the spatial performance of 243 undergraduates. Hierarchical regression analyses revealed that solution strategy accounted for significant, unique, and nonrcdundant amounts of performance variation for both mental rotation and orientation tasks, at a level roughly equivalent to the variation accounted for by sex. This was so even when strategy entered the regression equation after the variability due to sex, handedness, familial handedncss, reasoning ability, and cognitive habits had been removed. RESUME Dans unc scric d'etudes, trois strategics de solution de problcmcs spatiaux sont identifiees: une strategic de rotation mentale, une strategic de changement dc perspective et une strategic analytique. La fiabilite test-retest, la validitc et la generalisation d'unc mesure dc ccs strategics a etc ctablic. Cettc mesure, le questionnaire de strategie spatiale, fut ensuite ulilisee pour evaluer le role de la slragegie de solution dans la performance spatiale de 243 etudiants de premier cycle. Les analyses de regression hierarchiques ont revele que la solution de strategie expliquait une variation de performance significative, unique ct non-rcdondantc pour les taches de rotation et d'orientation, et cc a un niveau a peu pres equivalent a la variation obscrvec avec Ic scxe. Ceci ctait aussi vrai aprcs 1'elimination dc la variance due au sexe, a la dominance manucllc, a la dominance manucllc familialc, a la capacitc dc raisonnement et aux habitudes cognitives.People differ in how well they perform spatial tasks. Men score higher than women on a wide range of spatial problems (Maccoby & Jacklin, 1974; Sanders, Soares, & D'Aquila, 1982), but both within and between sexes this difference is moderated in a complex and not yet clearly delineated fashion by variables which include handedness, familial handedness, reasoning ability, and cognitive habits (Burnett, Lane,