Most agricultural areas are located in arid zones and drought is one of the most critical limiting factors of photosynthesis and crop yields, hence, treatments like methanol spraying can partly reduce damage from water stress. To investigate the effects of times and number of methanol spraying on some physiological characteristics of soybean under deficit irrigation, a factorial split-plot experiment based on a randomized complete block design with four replications was done at Karaj, Iran, in 2010. The first factor was drought stress in two levels (based on depletion of a1=40% and a2=70% of available soil moisture). The second factor was spraying times of methanol in two levels (in the morning at b1=8-10 AM and in the evening at b2=19-21 PM). Third factor was foliar application number of methanol with three levels (each c1=7, c2=14 and c3=21 days, Methanol spray was applied 5, 3 and 2 times during growth season of soybean, respectively). All treatments were sprayed with 21% (v/v) methanol concentration. 2 g lit-1 glycine was added to prepared solutions. In this study, grain yield, oil and protein percentage and yields, relative water content (RWC), leave chlorophyll content were determined. Results indicated a significant different (at 0.01probability level) between different levels of water stress on all evaluated traits and normal irrigation showed more favorable effects on measured parameters. Soybean grain yield under normal and deficit irrigation were 3187 and 1526 kg ha-1, respectively. However, no significant difference (P>0.05) was shown between different times of methanol spraying over studied characteristics. In contrast, measured traits were significantly affected by the number of methanol application. Moreover, results demonstrated that significant differences exists (p>0.05) between interaction effects a×b, a×c, b×c and a×b×c in all traits evaluated. In normal and deficit irrigation maximum grain yield was produced by methanol spraying every other week in the evening and every 7 days in the morning, respectively. Grain yield were positively correlated with relative water content (RWC) (r=+0.89**) and chlorophyll content (r=+0.94**). It seems by measuring these traits can predict yield at a certain period.