Trees and shrubs establishing in grasslands frequently occur in clumps as- sociated with increased soil moisture and N availability. We tested whether the presence of conspecifics or enhanced soil resources increased the growth and survivorship of a shrub (Symphoricarpos occidentalis) and a tree (Picea glauca) in the presence of the perennial grass (Bromus inermis). Rhizomes of Symphoricarpos and seedlings of Picea were trans- planted into plots either singly or with four nonspecific neighbors, and with grass neighbors present or removed. Half the plots received additional water, N, and shade to simulate a forest environment. Roots and shoots of transplants were harvested after two growing seasons. Soil moisture and available N were lowest at high transplant density, and highest in the simulated-forest environment and in treatments where grass was removed. Transplant survivorship was generally enhanced in plots where grasses were removed and in the simulated-forest environment. Picea survivorship was reduced by grasses in plots without resource manipulations but was unaffected by grasses in the simulated-forest environment, suggesting that the simulated-forest environment enhanced the ability of Picea to compete with grasses. Picea growth, however, was reduced by nonspecific neighbors in all cases. For Symphoricarpos, nonspecific neighbors significantly decreased growth rates with grasses absent but significantly increased growth rates with grasses present. Thus the direct effect of high density in monoculture was to reduce Symphoricarpos growth, but the indirect effect of high density in the presence of grass was to facilitate Symphoricarpos growth. In summary, nonspecific neighbors increased the growth rate of Symphoricarpos in competition with grass, and a simulated-forest environment enhanced the ability of Picea to survive in the presence of grasses. The results support the idea that contagious distributions of woody species invading grasslands enhance woody plant establishment.