To understand the nature and etiology of biases in geographical judgments, the authors asked people to estimate latitudes (Experiments 1 and 2) and longitudes (Experiments 3 and 4) of cities throughout the Old and New Worlds. They also examined how people's biased geographical judgments change after they receive accurate information ( "seeds" ) about actual locations. Location profiles constructed from the pre-and postseeding location estimates conveyed detailed information about the representations underlying geography knowledge, including the subjective positioning and subregionalization of regions within continents; differential seeding effects revealed between-region dependencies. The findings implicate an important role for conceptual knowledge and plausible-reasoning processes in tasks that use subjective geographical information.Geographical units like cities, provinces, countries, and continents are almost always irregular in shape and area; in their orientation relative to the cardinal points of the compass; and in their alignment relative to adjacent geographical units. Yet they also fit into a simple hierarchical scheme in which towns and cities are nested within provinces, provinces within countries, and countries within continents. It is obvious that people know about the spatial aspects of world geography as well as the hierarchical relations among geopolitical entities, and it is equally obvious that their knowledge is imperfect. What is less clear is how imperfect geographical knowledge is represented and how it is coordinated and weighed when used in making judgments.The tendency to normalize irregular shapes, combined with the tendency to develop a conceptual understanding of the hierarchical relations between geographical units, implies that geographical intuitions can be biased. In this article, we focus on understanding what geographical biases imply about how people encode, store, and use geographical knowledge in making judgments about locations. We take the position that most geographical biases result from plausible-reasoning processes that operate on complex beliefs about global geography and that few biases result from distortions of perceptually based representations. We conclude that the "mental map" metaphor is a misleading analogy for the representation of geographical knowledge and that the plausible-reasoning framework can account for much of the data in the subjective geography literature.We explored the nature of geographical biases by asking This research was supported by grants from the Natural Sciences and Engineering Research Council of Canada. We thank Ed Comell and Peter Dixon for thorough and thoughtful reading of a draft. We also thank Javier Movellan for suggesting that the seeding method could be used to study subjective geography.