The past several decades have seen a tremendous growth in the depth and breadth of our knowledge regarding social networks. While the majority of research to date has centered on the study of networks within small communities, the properties of networks in larger populations constitute a topic of increasing interest. Prior research has pointed to geography as a key factor in network formation, suggesting a spatial approach to the description and modeling of large-scale network structure. Further progress toward understanding the dynamics of large-scale social networks is hampered by the lack of data on the geographical structure of personal ties, and by the relative paucity of specialized methods for analyzing such data. The purpose of this project is to address these issues through two basic thrust areas: the collection of novel data on the spatial structure of large-scale social networks, and methodological research that will allow these data to contribute to our scientific understanding of social structure. The data to be collected will contain information about the spatial and demographic structure of personal ties at regional and national scales. These data will enable the estimation and evaluation of models for the effects of geography on network structure, and for the impact of social networks on neighborhood-level outcomes. The methodological research conducted as part of this project will enhance current network models — which characterize the probability of different types of social ties between individuals — by including the effects of spatial constraints, geographic heterogeneity, and demography. These data and methodological contributions will enhance fundamental knowledge about the complex ways that individuals interact with other through space. To perform these tasks, this project brings together an interdisciplinary team drawn from the fields of sociology, geography, and criminology; the team’s research will leverage current developments in network modeling and spatial statistics, as well as computational advances in the simulation of networks with large numbers of nodes. In addition to expanding our knowledge of the social ties which help knit communities together, the project will create new data and software tools that will be shared with other researchers and educators. The practical utility of the methodology and tools developed through this project will also be demonstrated with an application of spatial network modeling to the assessment and calibration of information dissemination systems for use during natural disasters and other large-scale emergency situations. By measuring and modeling the properties of large-scale social networks, this project will yield new insights into the social fabric in which people are all embedded.