The early detection and eradication of new invasive plants is often touted as the most cost-effective approach to invasive species management. However, this endeavor requires sufficient data infrastructure to track species across jurisdictional boundaries and relay information on new species’ locations. This talk details our experience in “horizon scanning” for future invaders in west-central North America by first compiling a simple species list of non-native plants using 1) non-native plant checklists curated by each state, and 2) digitized herbarium specimens. We found that assembling a list was hindered by tremendous variation in the quality, completeness and accessibility of checklists curated by individual states, which we quantified with a Data Quality/Accessibility Index. This index was positively correlated with the length of species checklists, suggesting that short checklists may not be a true representation of species richness in some states. For states with a high index, the length and composition of the species list was relatively stable regardless of which compilation method was used, whereas the two methods produced very different lists for states with poor indices. This study shows that we currently lack the infrastructure to track plant invasions in the west, even at coarse spatial resolutions. However, the presence of high-quality species lists in some states indicates that this problem is solvable with investments in digital infrastructure and taxonomic experts.