The gypsy moth has been gradually expanding its range in North America since its introduction approximately 150 years ago. One key question in this range expansion is the fate of isolated colonies that may be subject to Allee effects and/or demographic stochasticity. We used a novel spatial statistical tool, the Local Indicator of Spatial Autocorrelation, to identify isolated populations of gypsy moths that are sampled under the Slow-the-Spread project. We modeled the probability of a population persisting in a year given the population abundance in the pervious year and its relationship to a suite of ecological factors. A regional effect in the pattern of persistence was found such that populations in Wisconsin were more likely to persist than in other geographic regions of the project. The persistence of gypsy moths may be linked to recent studies that have shown a more rapid rate of spread in Wisconsin.