Nonnative perennial grass seedings for rehabilitating degraded lands and enhancing livestock forage are common across large areas in the western United States and other dryland regions. While these seedings can provide benefits in terms of perennial grass abundance and competition with invasive plant species, they also may reduce native plant abundance and diversity, with negative impacts on habitat for some wildlife species. We evaluated the effects of seeding a widely introduced nonnative perennial grass, crested wheatgrass (Agropyron cristatum), on native plant biodiversity in sagebrush steppe in eastern Oregon, focusing on differences in native forb (nongrass) understory species composition. We measured forb composition and other community characteristics at different scales in 17 sites across eastern Oregon with paired plots across seeding boundaries. We quantified differences associated with seeding overall. We also investigated the potential for seeding outcomes to relate to biotic interactions based on abundance of major functional groups, species traits related to competition, microsite features relating to competition, and soil texture. Finally, we tested the generality of our paired plot results across the study area by accounting for a broader range of management and environmental factors with a matching analysis using a large independent vegetation dataset. Paired plot comparisons suggested negative relationships between seeding and forb density. Biotic interactions with major functional groups were not strongly associated with forb community differences due to seeding, with the exception of a negative relationship between native annual forb abundance and increasing invasive annual grass cover. Few biotic interactions were related to seeding outcomes, though native annual forb abundance was negatively associated with higher invasive annual grass cover. Traits related to competition were not strongly related to species seeding response. Soil texture affected several forb community characteristics, including sometimes altering seeding outcomes. Increasing sand content was associated with lower perennial forb diversity and abundance, and annual forb richness decreased with higher sand content in seeded, but not unseeded, sites. The regional-level matching analysis, which controlled for the effects of management and environmental factors, generally supported the paired plot comparisons by showing that crested wheatgrass seeding was associated with lower invasive annual grass cover and lower total and perennial forb cover and richness. These results confirm earlier studies demonstrating lower invasive annual grass abundance in crested wheatgrass seedings and the effects of environmental variation, particularly soil type, on forb communities in crested wheatgrass seedings. There were few indications of strong aboveground competitive effects of crested wheatgrass on forbs, though inference was limited by the wide site-level variation in forb communities and the low abundance of many forb species, even in unseeded areas. Neither analysis was able to account for potential effects of degradation prior to seeding on forb communities. Additional research is needed to identify appropriate management strategies across environmental gradients, such as soil texture, to meet multiple goals such as controlling invasive annual grasses and enhancing diversity across the large areas historically seeded by crested wheatgrass and other nonnative perennial grasses.