This paper examines the factors that drive industry agglomeration, using coagglomeration patterns—how different industries tend to locate near each other—as a key empirical tool. The authors propose an index of coagglomeration and use data from the U.S. Census Bureau's Longitudinal Research Database from 1972 to 1997 to compute pairwise coagglomeration measurements for U.S. manufacturing industries. They assess the importance of three theories of industry agglomeration: (1) saving transport costs by proximity to input suppliers or final consumers, (2) allowing for labor market pooling, and (3) facilitating intellectual spillovers. The study finds that input-output dependencies are the most important factor, followed by labor pooling. The results suggest that industries are geographically concentrated due to a combination of these factors, with transport costs and labor market pooling playing significant roles. The paper also discusses the measurement of coagglomeration and the use of British input-output tables, employment patterns, and patent citations as instruments to address potential endogeneity issues.This paper examines the factors that drive industry agglomeration, using coagglomeration patterns—how different industries tend to locate near each other—as a key empirical tool. The authors propose an index of coagglomeration and use data from the U.S. Census Bureau's Longitudinal Research Database from 1972 to 1997 to compute pairwise coagglomeration measurements for U.S. manufacturing industries. They assess the importance of three theories of industry agglomeration: (1) saving transport costs by proximity to input suppliers or final consumers, (2) allowing for labor market pooling, and (3) facilitating intellectual spillovers. The study finds that input-output dependencies are the most important factor, followed by labor pooling. The results suggest that industries are geographically concentrated due to a combination of these factors, with transport costs and labor market pooling playing significant roles. The paper also discusses the measurement of coagglomeration and the use of British input-output tables, employment patterns, and patent citations as instruments to address potential endogeneity issues.