The ability to represent the spatial structure of the environment is critical for successful navigation. Extensive research using animal models has revealed the existence of specialized neurons that appear to code for spatial information in their firing patterns. However, little is known about which regions of the human brain support representations of large-scale space. To address this gap in the literature, we performed three functional magnetic resonance imaging (fMRI) experiments aimed at characterizing the representations of locations, headings, landmarks, and distances in a large environment for which our subjects had extensive real-world navigation experience: their college campus. We scanned University of Pennsylvania students while they made decisions about places on campus and then tested for spatial representations using multivoxel pattern analysis and fMRI adaptation. In Chapter 2, we tested for representations of the navigator's current location and heading, information necessary for self-localization. In Chapter 3, we tested whether these location and heading representations were consistent across perception and spatial imagery. Finally, in Chapter 4, we tested for representations of landmark identity and the distances between landmarks. Across the three experiments, we observed that specific regions of medial temporal and medial parietal cortex supported long-term memory representations of navigationally-relevant spatial information. These results serve to elucidate the functions of these regions and offer a framework for understanding the relationship between spatial representations in the medial temporal lobe and in high-level visual regions. We discuss our findings in the context of the broader spatial cognition literature, including implications for studies of both humans and animal models.