Understanding species-environment relationships is important to predict the spread of non-native species. Yellow toadflax (Linaria vulgaris Mill.) is an invasive perennial recently found in the Flat Tops Wilderness of the White River National Forest on the western slope of the Colorado Rocky Mountains. We hypothesized yellow toadflax occurrence could be predicted from easily measured site characteristics. We used logistic regression with stepwise selection to generate a model to predict yellow toadflax occurrence on a particular plot based on that site's physical characteristics. The experimental design was a paired-plot study in two locations using circular 1,018-m2 plots. Sixty-eight plots that did not contain yellow toadflax and 65 plots that contained yellow toadflax were sampled at the Ripple Creek site in 1999. In 2000, 54 non-toadflax plots and 55 toadflax-containing plots were sampled in the Marvine Creek site. Site characteristics sampled included: vegetation type; under-canopy light level; slope; aspect; soil properties; presence of disturbance, trails, and/or water; and total species richness. A model that correctly classified >90% of the 242 plots sampled included two vegetation type parameters, the presence of trails, and total species richness. Yellow toadflax is most often found in areas that were open-canopy sites, along trails, and with higher species diversity plots (>23 species). This approach can be used for other species in other areas to rapidly identify areas vulnerable to invasion. ?? 2007 Springer Science+Business Media, Inc.