Forest pest insects cause major socio-economic impacts, global losses of millions of dollars, and ecosystem changes. A key challenge for their management is tracing regional dispersal events critical to outbreak dynamics. We developed an integrated tracing framework for pest insects by combining isotope geolocation, ecological data, and atmospheric modeling, and applied this framework to the eastern spruce budworm moth (Choristoneura fumiferana), the most severe defoliator of the North American boreal forest, to trace outbreak dispersal events. We first generated a North American model of bioavailable sulfur isotope (δ34S) variation in space (isoscape) and then calibrated it to spruce budworm tissues of known origin. We then used an automated trap network with high temporal resolution to collect samples and identify potential immigration events of eastern spruce budworm to Nova Scotia, Canada. Finally, we traced the natal origin of these immigrants by sequentially integrating high-probability regions of origin derived from δ34S values and estimated migration routes derived from biologically constrained atmospheric transport models. We find that this integrated framework allows us to narrow down the region of pest origins, restricting it to a few possible locations and demonstrating long-distance dispersal of spruce budworm across ~400 km over the Gulf of St. Lawrence, Quebec. Our framework demonstrates that combining isotopic data with ecological indicators and atmospheric transport modeling offers improved resolution and understanding of insect dispersal ecology. This approach is transferable to trace other migratory insect species to address conservation, agriculture, and bio-surveillance needs in the context of global environmental change.