The susceptibility of a site to invasion by nonnative species depends on its current ecological features and its historical land use. Certain environments might be more conducive to an invasive plant’s success, and several recent studies have shown that former agricultural sites are more susceptible to invasion than sites that have been continuously wooded. We studied the invasive herb garlic mustard (Alliaria petiolata), in two regions with distinct ecological characteristics (the Connecticut River Valley and the Housatonic River Valley in Massachusetts), and two historical land uses (wooded versus cleared in 1830). A significantly (p less than10^-10) higher proportion of the random survey sites contained garlic mustard in the Housatonic River Valley than in the Connecticut River Valley, while there was only a marginally significant (p=.057) difference in garlic mustard presence among sites with different 1830 land use. Among sites that contained garlic mustard - both random survey sites and 9 previously known sites - ANOVA showed that land use history, but not ecoregion, was a significant determining factor in the amount of garlic mustard present, with bigger populations in sites that were open in 1830. Thus, the ecoregion appears to primarily determine the number of sites colonized, while land use history determines performance within colonized sites. This suggests that for garlic mustard, the role of 1830 land use is not in facilitating initial establishment, but in serving as a cause or indicator of environmental quality for proliferation of the invasion. It also suggests that mechanisms of dispersal may differ in the two regions. Further studies should be conducted to determine what aspects of ecology and historical land use are relevant the success of garlic mustard.