In highly invaded ecosystems, restoration of native plant communities is dependent upon reducing exotic species relative to native species. Even so, in monitoring, the native–exotic species richness ratio has been shown to be scale-dependent. Measurement at small spatial scales (<1 m2) can reveal a negative native–exotic richness relationship, where niche occupation may prevent invasion. Conversely, at larger scales, a positive correlation may exist, where environmental heterogeneity and equally favorable conditions may drive native–exotic relationships. Here, we compare slopes of native–exotic relationships across spatial scales in a prairie undergoing active restoration. The observed native–exotic richness ratios varied considerably over scales ranging from 1 to 1,000 m2, emphasizing the importance of choosing a measurement scale that is most pertinent to the treatment and ecological mechanism used to evaluate restoration success. Our native–exotic richness slopes were positive over all scales, but lower than would be expected in a random community assembly, suggesting the influence of niche-based competition. Correspondingly, our native–exotic cover slope was more negative than a null model; however, areas of frequent fire treatments showed a significant deviation from null only for richness, indicating that burning may enhance native–exotic competitive dynamics for number of species but not cover. The negative native–exotic cover relationships appear to be driven in this system mainly by exotic graminoids, across burn treatments and native functional groups, supporting the concept that frequent burning can alter the dominant competitive mechanism from coverage of these exotic grasses to an improved environment for germination and dispersal of more native species.