Shrubs seeded during rangeland restoration often fail to adequately establish. One factor differentiating successful and unsuccessful shrub seeding efforts is competition from grasses seeded with shrubs and weeds. Despite efforts to control these neighbors in experiments, their abundances often remained high, which suggests neighbors may have limited shrubs in much previous research. In the northern Great Plains, I evaluated seeded shrub survival and growth across a grass and weed biomass gradient. The treatment that held neighbor cover lowest provided the greatest shrub density and size, but density was low even in this treatment. Nevertheless, the data support previous studies indicating controlling neighbors can greatly benefit shrubs. Constraining neighbors increased shrub density ∼25 × (from about 0.005 to 0.13 m^-2). This large proportional effect could make the difference between shrub restoration success and failure when greater emergence occurs due to factors unrelated to competition, such as shrub seed rate and environmental conditions. Controlling neighbors is often necessary to establish shrubs, but currently, shrubs can fail to establish even where neighbors are well controlled. Additional research is needed to overcome barriers to establishment unrelated to competition.