No standardized, objective methodology exists for optimizing seeding rates when establishing herbaceous plant cover for pastures, hay fields, ecological restoration, or other revegetation activities. Seeding densities, fertilizer use, season of seeding, and the interaction of these treatments were tested using native plants on degraded sites in northern British Columbia, Canada. A mixture of 20% Achillea millefolium, 20% Carex aenea, 20% Elymus glaucus, 20% Festuca occidentalis, 16% Geum macrophyllum, and 4% Lupinus polyphyllus seed was applied at 0, 375, 750, 1,500, 3,000, and 6,000 pure live seed (PLS) per m2 in 2.5 × 2.5–m rototilled test plots, established in the fall and spring, with and without fertilizer. There was no significant difference in plant cover of sown species between fall seeding and spring seeding, and few treatment interactions were identified in the first 2 years after sowing. There was no significant difference in cover between seed densities of 3,000 and 6,000 PLS/m2 in the first year, nor among 1,500, 3,000, and 6,000 PLS/m2 treatments in the second year. Seed densities as low as 375 PLS/m2 produced year 2 plant cover equivalent to that observed at 3,000 PLS/m2 in year 1. Plots sown to seed densities less than or equal to 750 PLS/m2 generally exhibited an increase (infilling) in plant density from year 1 to year 2, whereas plots sown to seed densities greater than or equal to 1,500 PLS/m2 generally exhibited a decrease (density-dependent mortality) in plant density. These results imply a most efficient sowing density between 750 and 1,500 PLS/m2 (corresponding to 190–301 established plants.m−2 after two growing seasons). It is suggested that net changes in plant populations observed over a range of sowing densities are a robust and objective means of determining optimal sowing densities for the establishment of herbaceous perennials.