Tillage practices that maintain crop residues on the soil surface help reduce evaporation of soil water, which can benefit high water use crops such as maize (Zea mays L.). Management practices, climatic conditions, and soil type may affect how well a crop responds to surface residue. We conducted experiments with short season maize in 1994 and 1995 in Bushland, TX, USA, utilizing a rain shelter facility that has lysimeters containing monolithic cores of the Pullman (fine, mixed, thermic Torrertic Paleustolls), the Ulysses (fine-silty, mixed, mesic Aridic Haplustolls), and the Amarillo (fine-loamy, mixed, thermic Aridic Paleustalfs) soil series. In 1994, the treatments were a flat wheat (Triticum aestivum L.) straw and coconut (Cocus nucifera L.) fiber mulch of 4 Mg ha(-1) with infrequent irrigations totaling 25% and 75% of long-term average rainfall for the growing season (200 mm). The 1995 treatments were similar, but used a heavier mulch of 6.7 Mg ha(-1) and more frequent irrigations totaling 60% and 100% of long-term average rainfall. The mulch was applied at the 3-leaf growth stage. Mean potential grass reference evapotranspiration for the vegetative and reproductive growth stages in 1994 was 6.6 and 6.3 mm day(-1) respectively, and in 1995 it was 6.8 and 7 mm day(-1), respectively. The mulched and bare soil surface treatments used similar amounts of water in each year. In 1994, mulch did not affect yield, yield components, or leaf area index (LAI). No significant differences occurred in plant available water (PAW) between mulched and bare soil treatments from emergence through harvest. In 1995, mulch increased grain yield by 17%, aboveground biomass by 19%, and grain water use efficiency (WUE) by 14% compared with bare soil treatments. Mulched treatments also maintained significantly greater PAW compared with bare soil treatments until near anthesis and, after anthesis, LAI was significantly greater in the mulched treatments compared with the bare soil treatments. In 1995, mulch significantly increased grain yield and grain WUE of the maize crop in the Pullman soil, grain yield and biomass WUE of the crop in the Amarillo soil, and had no significant effect on the crop in the Ulysses soil compared with the bare soil treatments. The significant increase in water use efficiency in 1995 was the result of soil water being used for crop growth and yield rather than in evaporation of soil water. The more favorable soil water regime in 1995 compared with 1994 between the mulched and bare soil treatments was possibly due to the higher evaporative demand environment, the increase in mulch mass, and the increased irrigation frequency. This was especially important in soils where textural characteristics affected both rooting and soil water extraction by maize which limited its ability to tolerate water stress. (C) 1999 Published by Elsevier Science B.V.