Soil-borne seed pathogens are omnipresent but are often overlooked components of a community’s biotic resistance to plant naturalization and invasion. Using multi-year greenhouse experiments, we compared the seed mortality of single invasive, naturalized, and native grass species in sterilized and unsterilized soils collected from Pacific Northwest (USA) steppe and forest communities. Native Pseudoroegneria spicata displayed the greatest seed mortality, naturalized Secale cereale displayed intermediate seed mortality, and invasive Bromus tectorum was least affected by soil pathogens. Seed mortality across all three species was consistently greater in soils collected from steppe than soils collected from forest; seeds sown into sterilized steppe soil experienced half the overall seed mortality compared to seeds sown into unsterilized steppe soil. Soil sterilization did not affect grass seed mortality in forest soils. We conclude that (1) removing soil-borne pathogens with sterilization does increase native and non-native grass seed survival, and (2) soil-borne pathogens may influence whether an introduced species becomes invasive or naturalized within these Pacific Northwest communities as a result of differential seed survival. Soil-borne pathogens in these communities, however, have the greatest negative effect on the survival of native grass seeds, suggesting that the native microbial soil flora more effectively attack seeds of native plants than seeds of non-native species.