This study quantified herbaceous biomass responses to increases in honey mesquite (Prosopis glandulosa Torr.) cover on two soils from 1995 to 2001 in north central Texas. Vegetation was sampled randomly with levels of mesquite ranging from 0% to 100%. With no mesquite covering the silt loam soils of bottomland sites, peak herbaceous biomass averaged (±SE) 3 300 ± 210 kg · ha⁻¹ vs. 2 560 ± 190 kg · ha⁻¹ on clay loam soils of upland sites (P  =  0.001). A linear decline of 14 ± 2.5 kg · ha⁻¹ in herbaceous biomass occurred for each percent increase in mesquite cover (P  =  0.001). The slope of this decline was similar between soils (P  =  0.135). Herbaceous biomass with increasing mesquite cover varied between years (P  =  0.001) as did the slope of decline (P  =  0.001). Warm-season herbaceous biomass decreased linearly with increasing mesquite cover averaging a 73 ± 15% reduction at 100% mesquite cover (P  =  0.001) compared to 0% mesquite cover. Cool-season herbaceous biomass was similar between soils with no mesquite, 1 070 ± 144 kg · ha⁻¹ for silt loam vs. 930 ± 140 kg · ha⁻¹ for clay loam soils, but averaged 340 ± 174 kg · ha⁻¹ more on silt loam than on clay loam soils at 100% mesquite cover (P  =  0.004). Multiple regression analysis indicated that each centimeter of precipitation received from the previous October through the current September produced herbaceous biomass of 51 kg · ha⁻¹ on silt loam and 41 kg · ha⁻¹ on clay loam soils. Herbaceous biomass decreased proportionally with increasing mesquite cover up to 29 kg · ha⁻¹ at 100% mesquite cover for each centimeter of precipitation received from January through September. Increasing mesquite cover reduces livestock forage productivity and intensifies drought effects by increasing annual herbaceous biomass variability. From a forage production perspective there is little advantage to having mesquite present.