Global climate change is likely to drive up regional drought frequencies, aggravating China′s water shortage problem. Therefore, it is imperative to strengthen researches on crops adapting to the soil drought and agricultural water use efficiency. Soybean is an important source of high-quality protein and edible oil with relatively high water requirement. With an underdeveloped root system, it is the most drought-sensitive bean species. The number of pods per soybean plant is determined during the early stage of pod development. This stage is characterized by active cell division in the young ovules and is marked by rapid pod expansion. Drought stress occurring at this stage significantly increases the rate of pod abortion thus decreasing final seed yield. Therefore, this research focused on the effect of short-term soil water shortage during the pod-setting stage on leaf photosynthetic physioecology and final yield of the soybean, to provide a theoretical basis for water-saving cultivation and high-efficiently water using of soybean. It showed that leaf stomata appeared to be more sensitive to soil water content than leaf water content. On the second day of the experiment, leaf gs of the treatment group decreased to 24.9% of that of the CK (P＜0.01), but there was no statistically significant difference in leaf water potential (P＞005). Along with the water withholding, both leaf net photosynthetic rate and transpiration rate of the treatment group declined significantly during the drought, but ФPSⅡ only dropped on the third day of the experiment. On the third day after rewatering, net photosynthetic rate of the treatment group was 24.3% higher than that of the CK, showing a marked ultra-compensation effect after drought stress. Average seed dry weight, root/shoot ratio and harvest index of the treatment group were respectively 13.9% lower (P＜0.01), 35.3% higher (P＜0.05), and 17.4% lower (P＜0.05) than those of the CK. These results suggest that even a short-term drought during the pod-setting stage could influence later biomass distribution markedly.