Biological invasions pose a serious threat to biodiversity, but monitoring for invasive species is time consuming and costly. Understanding where species have the potential to invade enables land managers to focus monitoring efforts. In this paper, we compared two simple types of models to predict the potential distributions of three non-native invasive plants (Geranium robertianum, Hedera spp., and Ilex aquifolium) in the contiguous USA. We developed models based on the climatic requirements of the species as reported in the literature (literature-based) and simple climate envelope models based on the climate where the species already occur (observation-based). We then compared the results of these models with the current species distributions. Most models accurately predicted occurrences, but overall accuracy was often low because these species have not yet spread throughout their potential ranges. However, literature-based models for Geranium and observation-based models for Ilex illustrated potential problems with the methodology. Although neither model type produced accurate predictions in all cases, comparing the two methods with each other and with the current species distributions provided rough estimates of the potential habitat for each species. More importantly, this methodology raised specific questions for further research to increase our understanding of invasion patterns of these species. Although these types of models do not replace more rigorous modeling techniques, we suggest that this methodology can be an important early step in understanding the potential distributions of non-native species and can allow managers of natural areas to be aware of potential invaders and implement early detection.