The demand for restoration of degraded lands to diverse native habitat is growing, requiring efficient strategies for large-scale seeding and planting of native species. Restoration is often limited by low germination and establishment rates of native plants, so identifying the most effective seeding methods and rates may speed the restoration process. We tested three different methods of seeding (broadcasting, drilling, simulated hydroseeding) and five seeding rates (ranging from 0 to 1400 seeds/m²) to determine their efficacy in establishing three common species of Pacific Northwest prairies: Festuca roemeri, Eriophyllum lanatum, and Potentilla gracilis. We sowed seeds into six arrays at three western Washington sites on two dates (January and October) and monitored plant abundance for three years. We found that broadcast and simulated hydroseeding did not produce significantly different outcomes, suggesting that the extra resources required for hydromulching are not necessary. Additionally, broadcast seeding resulted in more consistent and reliable native plant establishment than seed drilling. Increasing seeding rates increased abundance, as expected, but species remained seed limited even at the highest seeding rates. Establishment varied considerably by site and seeding date. First-year establishment was positively correlated to third-year abundance, but this also varied greatly by site and species. Due to temporal and spatial variability in establishment, managers should evaluate treatments on individual sites and monitor results for several years after sowing.