Differences in soil microbial communities between ex-arable and undisturbed soils are often assumed to reflect long-term legacies of agricultural practices. Ex-arable soils, however, are commonly dominated by different plant species than undisturbed soils making it difficult to separate the importance of land-use and plant-growth legacies. In a system where non-native plants dominate ex-arable soils, we decoupled land-use (ex-arable, undisturbed) and plant-growth (native, non-native) effects on soil microbial communities using a factorial sampling design. Soils were removed from 14 sites that formed a 52-year chronosequence of agricultural abandonment. Microbial abundance and composition were measured using whole-soil phospholipid fatty acid analyses and microbial activity was measured in a subset of samples using sole-carbon-source utilization analyses. We found that both non-native-cultivated and ex-arable soils were independently associated with lower microbial abundance and diversity than native and undisturbed soils. We also found a correlation between microbial abundance and age-since-agricultural abandonment in ex-arable/non-native-cultivated soils suggesting that non-native plant effects accumulate over time. Microbial activity was consistent with microbial abundance; microbial communities in non-native-cultivated, ex-arable soils were slow to respire most carbon sources. Our data suggests that agricultural practices create soil conditions that favor non-native plant growth and non-native plants maintain these conditions. Potential mechanisms explaining how non-natives create soils with small microbial communities and how small microbial communities may benefit non-natives are discussed.