A major challenge to addressing biological invasions is that the need for emergency responses often precludes opportunities to analyze the dynamics between initial establishment and population eruption. Thus, a broader understanding of underlying processes and management opportunities is often lacking. We examined the effects of habitat structure and natural enemies on recently established preeruptive gypsy moth, Lymantria dispar L., populations over 4 yr in northeastern Wisconsin. Forty-five sites were established across a range of habitat structures in oak-dominated northern hardwood forests. The number of egg masses was positively related to percent composition of oaks and other favored species. Other life stages were not related to habitat structure variables. Abundance of each life stage can predict the subsequent life stage with variable degrees of accuracy, but male moth densities were only a weak predictor of egg mass or larval densities the following year. The parasitic fly Compsilura concinnata (Meigen), an introduced generalist with deleterious nontarget effects, caused the highest mortality to larvae. The specialist pathogens Entomophaga maimaiga and nucleopolyhedrosis virus were widely distributed but caused less mortality than reported in the northeastern United States, where gypsy moth has been established much longer. Small mammals are the major predators of pupae as elsewhere, but invertebrates seem less important along the western than southern advancing front of gypsy moth. Overall habitat structure did not influence natural enemy populations. These results suggest that the pre-eruptive phase is distinct from the pre-establishment phase by high mating success and from the eruptive phase by the prominent role of generalist natural enemies. Improved understanding of these dynamics can help guide silvicultural and biocontrol strategies in newly invaded regions of the Midwest and provide general insight into invasive forest defoliators.