This project explains variations in party-association linkages among Mexican states in competitive and noncompetitive electoral environments. The subnational comparison includes two states governed by each of Mexico’s three major parties, one state where the party is electorally dominant and another where elections are competitive. In each of these six states, I study associations from both the traditional popular sectors (e.g., organized labor, peasant associations) and the new popular sectors (e.g., neighborhood associations, indigenous movements). Party-association linkage outcomes are explained by variations in electoral competition and national party organizations. Dominant parties in noncompetitive environments will tend to form stable material-based linkages, while parties in competitive electoral environments are hypothesized to form a combination of stable policy-based linkages and unstable material-based linkages. The linkage legacies and the levels of decentralization of national party organizations also influence whether parties on the state level form linkages with associations representing the traditional popular sectors or the new popular sectors. The subnational comparative method, large-n statistical analysis, and ethnography are combined to draw reliable causal inferences. The subnational comparison of six Mexican states allows me to observe variation on hypothesized causes of linkages while controlling for national-level factors. Through statistical analysis, using original survey data of associations and data from election results and budgets for state-administered social programs, I test the relationships among my causal variables on a larger sample. Ethnographic studies of a neighborhood association and a union of social security employees in Mexico City allow me to flesh out the cultural and organizational mechanisms that lead to the construction and reproduction of party-association linkages.