Four months before planting a crop of maize, a study was commenced on a well-drained clay loam soil to test the effect of tillage and surface maize residue on soil water storage. Tillage main plots were split into bare (residue removed) and residue-covered sub-plots. At commencement, all plots were irrigated to field capacity to a depth of 600 mm. Initially, large positive differences in soil water storage developed due to residue cover in the tilled treatments (chisel and disc respectively), but this advantage tended to diminish during the prolonged dry period. Tillage accelerated drying of the surface soils, and the resultant dry layers restricted movement of subsoil water to the surface, causing lower rates of evaporation during long, dry periods. The relatively faster rates of water loss from the covered and bare direct-drill (DD) sub-plots during the dry winter was ascribed to the continuous micropores linking the subsoil and soil surface, which promoted supply of subsoil water to the surface. Residue cover was most effective in reducing stage-one evaporation, but when the rainless period exceeded about 2 weeks, the moister soils under residue resulted in rates of evaporation which equalled and then exceeded those of bare soils. Surface residues were most effective in increasing soil-water storage when rainfall frequency was high just prior to planting, with greater storage where residue cover was higher. Among the bare sub-plots, DD had the driest profiles at planting, whereas the tilled profiles were drier where soil disturbance had been greater.