We examined biotic soil crust cover and composition at nine shrub-steppe sites in central and eastern Oregon, U.S.A. One pair of livestock-grazed and excluded transects was established at each site. Data were collected on the cover of biotic soil crust and vascular plant species, soil surface pH and electrical conductivity, and other environmental variables.Using gradient analysis, we found that differences in community composition among sites were most strongly related to soil pH, electrical conductivity (EC), and Calcareous Index Value (CIV; a scale representing the relative calcium carbonate content of soils). Other important variables included precipitation, elevation, aspect, and temperature. We found total crust cover to be highest at sites with lower pH, EC, and CIV. Dominant species differed markedly between the more calcareous sites with higher pH, and the less calcareous, lower pH sites. Livestock exclusion was not an important gradient in the ordination of these data, being overshadowed by the strong soil chemistry and climate gradients. However, overall community composition of soil crust species was different between grazed and long-ungrazed sites (p = 0.02, Blocked Multi-Response Permutation Procedure). Comparison of grazed and long-ungrazed sites revealed lower cover of biotic crusts, nitrogen-fixing lichens, crust-dominated soil surface roughness, and lower species richness in the grazed transects. There was more bare ground in the grazed transects, on average (p ≤ 0.02 for all, two-tailed paired t-tests). Our results suggested that total bunchgrass cover was higher within exclosures, but conclusive evidence was lacking (p = 0.1, two-tailed paired t-test). Vascular plant composition, cover, richness, shrub cover, electrical conductivity, and pH were not different between the grazed and livestock-excluded transects. Thus, livestock-related reductions in cover and richness of biotic soil crusts were apparent while significant impacts to vascular plants were not obvious. We conclude that 1) biotic soil crusts are sensitive indicators of disturbance and 2) there are strong compositional differences in shrub steppe crust communities of Oregon, which are correlated with regional soil and climate gradients.