This study tested two contradicting predictions: preschool children’s inequity aversion is a deliberative cognitive-controlled act or that it is an automatic act. The ultimatum game models social exchange in situations in which the rational motive to maximize gains conflicts with fairness considerations. In this study, Ultimatum Game variant was adopted to present allocation proposals by third-party, and we aimed at explaining the subjects’ decisions more accurately, because allocating intension was removed from the proposal. Mixed experimental design was 2(cognition resource)×2(proposal type) ×3(age group), then cognition resource and proposal type were between subject variables, to explore how preschool children’s disadvantageous and advantageous inequity aversion depend upon cognition resources. Operating cognitive resources by depletion task----Day-Night stroop task, and non-depletion task was designed when subjects should accomplish UG task with rich cognition resources. There were 12 UG trials, including equity proposal and inequity proposal 2 types. Before each trial, 1to 4 depletion tasks would be presented firstly in order to deplete cognition resources fully. All tasks were presented with computer, and depletion / non-depletion task and UG task were presented alternatively. Subject’s decision (reject or accept) was coded. We found that, （1）4-to-6-year-olds showed disadvantageous inequity aversion and it increased as they got older (F（2,101）=18.522, p<.001; M=3.33, SD=1.30; M=2.70, SD=1.40; M=1.25, SD=1.67). On the contrary, advantageous inequity aversion did not increase along with growing age and it expressed later. There was imbalance in development of these two kinds of inequity aversion (F（1,101）=76.557, p<.001, η2 =.431). (2) 4-to-6-year-old preschool children’s two inequity aversions were affected by available cognitive resources. If there were less cognitive resources, 4-to-6-year-olds would be more sensitive to the self interest and show low frequency of disadvantageous inequity aversion (F（1,103）=20.640, p<.001，η2 =.167; M=2.49, SD=1.68; M=2.97, SD=1.46) and advantageous inequity aversion (F（1,103）=21.558, p<.001, η2 =.173; M=3.63, SD=.70; M=3.87, SD=.42), and tended to accept unfair allocation proposal more easily. This illustrated that children’s inequity aversion was rational behavior controlled by cognition. (3) When cognitive resources were insufficient, compared with advantageous inequity aversion, the degree children’s disadvantageous inequity aversion reduced was higher (F（1,101）=5.973, p<.05, η2 =.056; M=.50, SD=.10; M=.24, SD=.05). This research revealed that 4-to-6-year-old preschool children’s inequity aversion is a deliberative cognitive-controlled act dependent upon available cognition resources.