A system which imposes a range of water stress levels was developed and evaluated. Water stress was controlled by employing a screen to suspend roots above a water column of known height. Levels of water stress were imposed by changing water column height and/or hydraulic conductivity of the medium in the column. The system was evaluated in a series of growth chamber experiments in which sunflowers (Helianthus annuus L. cv NK894) were given three levels of water availability for a period of 3 weeks. Third leaf midday water potentials at the end of the trials ranged from -0.73 +/- 0.04 to -2.35 +/- 0.17 megapascals in waterstressed plants compared to -0.40 +/- 0.02 megapascals for control plants. Repetition of experiments showed no statistical differences in leaf water potentials, plant leaf areas, or plastochron indices between trials. During the experiments, the severity and pattern of water stress developments was related to both water column height and conductivity of the medium. Control plants exhibited normal diurnal water relations and transpirational behavior. Use of this system avoids many problems associated with other techniques and provides a means for subjecting plants to reproducible water stress levels for extended periods of time.