Images, such as photographs and diagrams, play an important role in the teaching and learning of science. To optimize student learning, educational science images should be designed to facilitate the cognitive processes relevant to comprehension. One such process is comparison, which involves aligning multiple representations on the basis of their common relational structure. This structural alignment process can be facilitated by cognitive supports that are inherent to an image, including its spatial layout. Yet, little is known about the extent to which students must engage in comparison to learn from science images, and whether widely-used educational materials are conducive to structural alignment. To address these issues, we sampled multiple chapters from each of three popular U.S. middle school life science textbooks. We coded each image for the presence of prompts for comparison using cues within the images and surrounding text. For each image that prompted comparison, we coded whether its layout facilitated relevant structural alignment (direct placement of matched pairs) or obscured alignment (impeded placement). Overall, we found that comparisons were prompted for more than a third of the images. However, fewer than half of the images that required comparison had a spatial layout that provided strong support for comparison—that is, direct placement of matched objects/parts. We propose that, in concert with other cognitive supports for learning from multiple representations, spatial supports for comparison could be applied broadly to increase the effectiveness of educational science images.