Conservation of endangered species necessitates a full appreciation of the ecological processes affecting the regulation, limitation, and persistence of populations. These processes are influenced by birth, death, and dispersal events, and characterizing them requires careful accounting of both the deterministic and stochastic processes operating at both local and regional population levels. We combined ecological theory and observations on Allee effects by linking mathematical analysis and the spatial and temporal population dynamics patterns of a highly endangered butterfly, the high brown fritillary, Argynnis adippe. Our theoretical analysis showed that the role of density-dependent feedbacks in the presence of local immigration can influence the strength of Allee effects. Linking this theory to the analysis of the population data revealed strong evidence for both negative density dependence and Allee effects at the landscape or regional scale. These regional dynamics are predicted to be highly influenced by immigration. Using a Bayesian state-space approach, we characterized the local-scale births, deaths, and dispersal effects together with measurement and process uncertainty in the metapopulation. Some form of an Allee effect influenced almost three-quarters of these local populations. Our joint analysis of the deterministic and stochastic dynamics suggests that a conservation priority for this species would be to increase resource availability in currently occupied and, more importantly, in unoccupied sites.