We describe methods of coding and analyzing historical and archaeological data for comparative analysis of premodern cities. As part of a larger study of spatial access to urban services, we identify eight relevant contextual domains and define variables for each domain. Information from publications on each city is assembled, and the data are coded independently by three scholars and checked for agreement. To date, we have completed contextual coding for 15 cities. Here, we focus on methods to analyze relationships among variables within contextual domains using two example domains—Class Mobility and Governance. Key methodological points involve the problem of missing data, multiple tests with an appropriate correction, and the importance of variation among cases for effective analysis of a domain. Our interpretation of preliminary findings highlights a degree of independence between two different arenas of social life that may relate to collective action. Documentation of our procedures contributes to a growing body of systematic, comparative, cross-cultural analyses of mid-size samples. This study is distinctive in its focus on cities rather than cultures, societies, or polities.