(1) Data collected from coordinated comparative studies on the structure and function of ten central and western United States grasslands are presented and analysed. The ten study areas encompassed six major grassland types: mountain grassland, northwest bunchgrass, mixed-grass prairie, shortgrass prairie, tallgrass prairie and desert grassland. Experimental design at each location included two replications of ungrazed (i.e. long-term absence of grazing by large domestic herbivores) and grazed treatments. (2) Structural characteristics studied included quantities of cool- and warm-season plants at each site, and magnitude and seasonality of primary producer compartments. (3) The environmental variables measured at each site were precipitation, temperature and solar radiation. These variables were calculated as both annual and growing-season values, along with annual and growing-season potential and actual evapotranspiration. (4) The seasonal mean live biomass averaged 94 g m -2 for the ten grasslands, with a mean of 98 g m -2 on ungrazed grasslands and 89 g m -2 for grazed grasslands. Ungrazed grasslands averaged about 160 g m -2 of recent and old dead plant material, compared to approximately 70 g m -2 on grazed grasslands. Ungrazed grasslands had about 50 and 150% more recent and old dead material, respectively, than grazed grasslands. Generally, grazing in previous seasons decreased the standing crop of dead material across the grasslands. (5) The mean total shoot standing crop averaged 203 g m -2 for all grasslands across all treatments; ungrazed grasslands averaged 245 g m -2 , compared to 161 g m -2 for grazed grasslands. The magnitude of litter corresponded in general to the amount of total shoot standing crop. (6) Generally, the shortgrass prairie had the maximum amount of crown material, followed by the mixed-grass prairie, with the tallgrass prairie least. Crown material averaged c. 275 g m -2 for the mixed- and shortgrass prairie, compared to c. 200 g m -2 for the tallgrass prairie. Although grazing did not significantly influence the amount of crown material, crown material tended to increase under grazing in mixed-grass and shortgrass prairies. (7) Root biomass ranged from 156 g m -2 in the desert grassland to as much as 2000 g m -2 in the mixed-grass prairie, while other grasslands were intermediate between these two extremes. (8) There was an inverse relationship between root biomass and dynamics and mean annual temperature. Grazing generally resulted in an increase in the root/shoot ratios, particularly on the cooler grasslands.