Six open-top chambers were installed on the shortgrass steppe in north-eastern Colorado, USA from late March until mid-October in 1997 and 1998 to evaluate how this grassland will be affected by rising atmospheric CO2. Three chambers were maintained at current CO2 concentration (ambient treatment), three at twice ambient CO2, or approximately 720 μmol mol−1 (elevated treatment), and three nonchambered plots served as controls. Above-ground phytomass was measured in summer and autumn during each growing season, soil water was monitored weekly, and leaf photosynthesis, conductance and water potential were measured periodically on important C3 and C4 grasses. Mid-season and seasonal above-ground productivity were enhanced from 26 to 47% at elevated CO2, with no differences in the relative responses of C3/C4 grasses or forbs. Annual above-ground phytomass accrual was greater on plots which were defoliated once in mid-summer compared to plots which were not defoliated during the growing season, but there was no interactive effect of defoliation and CO2 on growth. Leaf photosynthesis was often greater in Pascopyrum smithii (C3) and Bouteloua gracilis (C4) plants in the elevated chambers, due in large part to higher soil water contents and leaf water potentials. Persistent downward photosynthetic acclimation in P. smithii leaves prevented large photosynthetic enhancement for elevated CO2-grown plants. Shoot N concentrations tended to be lower in grasses under elevated CO2, but only Stipa comata (C3) plants exhibited significant reductions in N under elevated compared to ambient CO2 chambers. Despite chamber warming of 2.6 °C and apparent drier chamber conditions compared to unchambered controls, above-ground production in all chambers was always greater than in unchambered plots. Collectively, these results suggest increased productivity of the shortgrass steppe in future warmer, CO2 enriched environments.