Relationships involving fire and perennial grasses are controversial in Chihuahuan Desert grasslands of southern New Mexico, USA. Research suggests that fire delays the resprouting of perennial grasses well after two growing seasons. However, such results are confounded by livestock grazing, soil erosion, and drought. Additionally, post-fire grass responses may depend on initial clone size. We evaluated the effects of fire, grazing, and clone size on Bouteloua eriopoda (black grama) in southern New Mexico grasslands. Four 2-ha plots were established in each of four sites. Fire and grazing were applied or not applied in 1999 such that four treatment combinations were assigned randomly to plots within each site. Within each plot, small (0-10 cm(2) basal area), medium (10-30 cm(2)), and large ( > 30 cm(2)) clones were initially mapped in five 0.91-m(2) quadrats where grass attributes and litter cover were evaluated before and at the end of two growing seasons following fire. Maximum fire temperature was also measured. At a population level, canopy and litter cover were each approximately 50% less in burned than unburned areas. However, compared to initial levels, canopy height had increased by 10% at the end of the study, regardless of fire. At a clonal level, basal cover reductions were attributed mostly to large clones that survived fire. Smaller clone densities had decreased by as much as 19% in burned compared to unburned areas, and fire reduced the basal cover of medium clones. Basal and canopy cover, recruitment, and clone basal area decreased with increased fire temperatures. Almost all responses were independent of grazing, and interactive effects of grazing and fire were not detected. Fire did not kill all perennial grass clones, regardless of size. However, rapid responses were likely influenced by above-average precipitation after fire. Future studies in desert grasslands should examine how perennial grass dynamics are affected by fire, precipitation patterns, and interactions with grazing.