Studies on the resistance of communities to plant invasions at different spatial scales have yielded contradictory results that have been attributed to scale-dependent factors. Some of these studies argue either for or against Elton's notion of biotic resistance against invasions through diversity. We studied the correlation between alien weeds and native species, dividing the latter group into weedy and non-weedy species, integrating various factors that influence diversity into an analysis on the scale of the federal states of Mexico. The resulting multiple-regression models for native and alien weed species are robust (adjusted R2 = 0.87 and R2 = 0.69, respectively) and show a strong partial correlation of the number of weed species (native and alien) with the number of non-weed native species. These results agree with studies showing a positive correlation between the number of native and alien species on larger scales. Both models also include human population density as an important predictor variable, but this is more important for alien weeds (β = 0.62) than for native weeds (β = 0.32). In the regression model for native weed species richness, the non-cultivated (fallow) area (β = 0.24) correlated positively with native weed richness. In the model for alien weed species richness, the native weed species richness was an important variable (β = −0.51), showing a negative partial correlation (rpart = −0.4). This result is consistent with Elton's biotic resistance hypothesis, suggesting that biotic resistance is scale independent but that this may be masked by other factors that influence the diversity of both weeds and non-weeds.