Water stress has been shown to alter protein metabolism in vegetative tissues of many plant species (2,15). Among the documented effects of water stress are: decreased tissue protein content (5, 10), decreased incorporation of amino acids into protein (3, 9), decreased nitrate reductase activity (19), increased ribonuclease activity (8, 12), dissociation of polyribosomes (13,14), and altered mRNA transcription and translation (16). Changes in the pattern of protein synthesis during water stress have been observed in double-labeling experiments (6, 7) indicating that the effect is selective. This raises the possibility that water stress may alter the nutritional quality of high protein legume crops such as soybean, the mature seeds of which are 40 to 45% protein on a dry weight basis. However, little is known about the effect of water stress on storage protein synthesis and accumulation in developing legume seeds.In this paper we report that protein accumulation in developing ovules of two soybean varieties is not prevented and can actually be increased by moderate or severe (leaves senesced) water stress. We also present evidence for stress-induced foliar nitrogen mobilization which appears to account for the maintenance of normal or supranormal seed protein accumulation. Greensboro, NC) at a concentration of 12.5 mg/l. KNO3 was the major N source in this solution. Young plants were grown in half-strength nutrient solution for the first 2 weeks after transfer into the hydroponic system and full strength solution thereafter. The pH of the solution was maintained between 6.2 and 6.5 and the solution was replaced twice a week. Aeration was provided by bubbling compressed air through the nutrient solution. Newly opened flowers were tagged daily with color-coded thread for approximately 1 month to provide a precise record of seed age at harvest.Administration of Water Stress. Stress was applied stepwise by addition ofPEG-6000 (Carbowax 6000, Union Carbide Corp.) to the nutrient solution in 2.5% (w/v) increments every 2 d to a final level of 12.5% PEG (about -9.2 bars). The PEG concentration was held at 12.5% for 10 to 13 d. For the severe stress experiment in the summer of 1981, the plants were grown in 2.5-L pots, and transpired solution was replaced with deionized H20 every evening. During the day the solution level was allowed to drop due to transpiration thereby decreasing the water potential to -15 to -20 bars. For the moderate stress experiment in the summer and fall of 1982, the plants were grown in 5-L pots such that transpiration had little effect on the solution level and the water potential remained between -9 and -12 bars. At the end of the stress period, pods were harvested, sorted by age, shelled, and the seeds frozen in liquid N and stored at -30°C.