Community dynamics and dominance on cheatgrass (Bromus tectorum L.) and yellow starthistle (Centaurea solstitialis L.) infested rangeland appear to be influenced by resource acquisition rates and duration of growth. Objectives were to determine the effects of densities, proportions, and soil depth on the growth rate and duration of growth of these species. In 6 field experiments isolated individuals, monocultures (100, 1,000, 10,000 plants m-2), and mixtures (same densities arranged factorially) were grown with unrestricted and restricted (0.2- and 0.5-m) soil depths. Shoot weights were determined on 12-day intervals beginning on day 24 and ending on day 72 for plants grown with restricted soil depth and day 96 (cheatgrass) and day 108 yellow starthistle) for plants grown in unrestricted soil. Quadratic growth curves were fit for each replication for plants grown in isolation. Linear and quadratic models season. We tested whether blue grama [Bouteloua gracilis (HBK.) Lag.] could be established in crested wheatgrass fields to increase plant diversity and grazing season length. Three seeding methods (unseeded, seed broadcast after rototilling, or seed drilled) and 2 herbicide treatments (none or glyphosate [N (phosphonomethyl)-glycine] sprayed at 1.1 kg a.i. ha-1) were applied in a complete factorial design to 3 X 10 m plots in 1994 in two 50 year-old crested wheatgrass stands. Blue grama established in seeded plots, but the effectiveness of seeding methods varied between sites. At both sites, herbicide control of crested wheatgrass greatly promoted the establishment of blue grama. Crested wheatgrass biomass and cover were reduced by rototilling and by spraying. The results suggest that the establishment of blue grama in crested wheatgrass stands is possible if seed is added and competition from crested wheatgrass is controlled.