This paper identifies subnational peripheries as regions with historical trajectories and social formations that differ from those the majority of state’s territory, and that present distinct inferential challenges to within-country comparative designs using the subnational comparative method (Snyder 2001). These challenges are directly interpretable using standard concepts in quantitative and qualitative political methodology: unit heterogeneity, unobserved heterogeneity, complex interactive causes, small-n problems, and nonignorable missing data. I illustrate the stakes of the argument for current practice using three questions for which the subnational comparative method is ideally suited—local public goods and economic development, identity and insurgency, and ethnic voting—in the context of six Southeast Asian countries: Indonesia, Malaysia, Myanmar, the Philippines, Thailand, and Vietnam. My discussion clarifies exactly how subnational peripheries complicate causal inference through within-country comparisons, and shows how many of the solutions to the problems raised by subnational peripheries frequently change the causal parameter of interest, or alternatively, the population being studied. These findings have important implications for current practice.