The generality of the water-limited-nitrogen-regulated plant growth hypothesis for desert ecosystems was tested by applying water, nitrogen and sulfur amendments in a complete factorial design to three Chihuahuan Desert ecosystems. Water was added every 2 weeks at a rate to about double the long-term annual precipitation average of 225 mm and nitrogen and sulfur were added once in May 1983 (10 g/m2). Living-canopy area of all herbaceous plants was measured monthly (from April 1983 to December 1984). Since plants can respond to water and nutrients either by establishing more, but smaller, plants or by growing fewer, but larger, plants, density and plant size were measured (once, in me-April 1984). Only C3 winter-ephermerals significantly changed in cover, density or plant size in response to the amendments; no significant responses were observed in C4 summer-ephemerals or perennial grasses and forbs. Species that significantly increased in density did not significantly increase in size and vice versa. Water tended to limit the number (richness) and density of species, whereas nitrogen regulated plant growth (size). Sulfur generally reduced soil pH and species density. This study demonstrated that these general plant responses to increased water and nutrients were highly species and site specific.