Population synchrony, or concurrent fluctuations in population density across a large geographical area, is a common phenomenon among forest insect populations. In many cases, the cause of synchrony is unknown. Simple models are very useful for making predictions for when population synchrony is expected to occur, but data often appear to violate those predictions. For instance, chaotic population dynamics and limited dispersal may be common among synchronous forest defoliators and both of these factors should theoretically make synchrony unlikely. Using the North American gypsy moth (Lymantria dispar) as a case study, we use mechanistic population models to explore how local population dynamics, interspecific interactions, environmental stochasticity, and dispersal can interact to promote long-distance synchrony. We find that the mechanistic models can explain instances of gypsy moth synchrony that are not predicted by simpler models.