There are a growing number of contrasting theories about plant invasions, but most are only weakly supported by small-scale field experiments, observational studies, and mathematical models. Among the most contentious theories is that species-rich habitats should be less vulnerable to plant invasion than species-poor sites, stemming from earlier theories that competition is a major force in structuring plant communities. Early ecologists such as Charles Darwin (1859) and Charles Elton (1958) suggested that a lack of intense interspecific competition on islands made these low-diversity habitats vulnerable to invasion. Small-scale field experiments have supported and contradicted this theory, as have various mathematical models. In contrast, many large-scale observational studies and detailed vegetation surveys in continental areas often report that species-rich areas are more heavily invaded than species-poor areas, but there are exceptions here as well. In this article, I show how these seemingly contrasting patterns converge once appropriate spatial and temporal scales are considered in complex natural environments. I suggest ways in which small-scale experiments, mathematical models, and large- scale observational studies can be improved and better integrated to advance a theoretically based understanding of plant invasions.