|Abstract:||This thesis was undertaken for the purpose of investigating the relationship between the cost characteristics of counties throughout the United States in 1860 and 1870 and the likelihood of the presence of large-scale industrial enterprise. I have attempted to redirect the analysis of manufacturing development during the American industrial revolution away from strictly regional data aggregation to a more narrow geographic focus--the county. Specifically, I utilized the Bateman-Weiss large firm samples for 1860 and 1870, as well as published census data and map information to answer the following question: to what extent did rail mileage, water access, capital market "potential" and the presence of an urban site at the previous census, influence the probability that any given county was the site of a large firm? The qualitative nature of this model--the dependent variable, presence of a large firm or not, is 0-1--necessitated the use of the logit model to obtain estimates of how small changes in the independent variables changed the probability of large-firm activity. My findings revealed that strong regional differences in locational behavior did exist during this period, although large-scale urbanization (cities of 20,000 or more inhabitants), consistently contributed the most to a county's suitability as an industrial site. Overall, the model predicted industrial counties best in the East. The West and South were quite similar in 1860; however, during the following decade southern counties were still highly reliant on rail transport and small-scale urbanization to attract large producers. It was necessary to conclude that individual factors were less important in determining the desirability of potential industrial sites. Rather, the package of benefits offered by urban areas--external scale economies, improved input and output market conditions, and transport nodes--was crucial to the birth and development of manufacturing enterprise in 1860 and 1870.