This study examined differences between two groups of students' spatial-scientific reasoning from pre to post implementation of an Earth/Space unit. Using a quasi-experimental design, researchers explored how instructional method and gender affected learning. Treatment teachers employed an integrated STEM curriculum while the control teacher implemented her regular Earth/Space unit. The Geometric Spatial Assessment (GSA), the Purdue-Spatial Visualization Rotation Test, and the Lunar Phases Concept Inventory (LPCI) were used to assess learning. Experimental groups made gains on periodicity LPCI domains while the control made gains on geometric spatial visualization LPCI domains. Only females made gains on GSA items. This is the first quasi-experimental study to examine students' spatial reasoning as they participate in Earth/Space units and to discover gender's role in this spatial development. [For the complete proceedings, see ED584829.]