While recent studies have shed light on the relationship between language and decision, less is known about diglossia and social decision-making, especially regarding empirical evidence. The present study was designed to investigate the effect of language type (dialect and mandarin) on diglossia speakers’ social decision-making by two experiments using within-participants design. The purpose of the first experiment was to determine whether dialect has an impact on the diglossia speakers’ group decision making by using adapted public good game (PGG), including interpersonal trust, cooperation and emotion level. Experiment 2 used ultimatum game (UG) to investigate the effect of dialect on individual’s social decision. Participants were undergraduates whose hometowns were Qingdao or Weifang in Shandong province and started to speak dialect from birth. All the recordings were recorded by proficient diglossia speakers and edited by Adobe Audition CS5.5. All the stimuli were presented on a 14-inch Lenovo computer by E-prime 2.0. In experiment 1, participants were told to play a game with other 3 undergraduates performed by our experimenters in separate room. Participants first completed PANAS questionnaire. Then we used physiological recorder MP150 to measure eletrocardio and skin conductance activity lasted for 5 minutes as baseline. Each subject has 30 yuan, and they can secretly choose how many of their private tokens to put into a public pot. The tokens in this pot are multiplied by two and this "public good" payoff is evenly divided among players. Each subject also keeps the tokens they do not contribute. Before they put tokens into the pot, they has 15s to express their opinions. Participants were told that 4 participants spoke in random but actually the procedure was fixed and the participant spoke at last in this way he may find that all the other participants spoke in the same dialect as him. After speaking stage, participant need to complete 4 ratings, including PANAS, trust rating, cooperation rating, level of similarity in dialect with other three undergraduates. Experiment 2 was 2 (language: dialect and mandarin)×4 (distribution plan: ￥5-￥5, ￥3-￥7, ￥2-￥8, ￥1-￥9) design. There was a total of 32 trails in formal experiment. In a trial, a black fixation was presented for 800ms first. Then, the distribution plan was played in two levels randomly in 5000ms. Next, the responders made their decisions whether they accepted and got the responding money or refused and neither of the two individuals received any money. After that was a feedback for 1500ms. At last, a blank screen was presented in 500ms. The results indicated that: (1) In PGG, participants expressed higher interpersonal trust (p = .044), cooperation (p = .017), positive emotion variation (p < .001), heart rate variation (p = .008) in dialect level than mandarin level. In mandarin level, the trust score correlated positively with cooperation score (p < .001) and negatively with negative emotion (p = .004). Furthermore, the trust score fully mediated the relationship between negative emotion and cooperation score. (2) In UG, the more equitable the proposals were, the higher the acceptance rate was. Participants had significantly higher acceptance rate of the distribution plans spoken in dialect than in mandarin, p = .03. These findings highlighted on the influence of dialect on people’s cognitive, emotion and decision-making, providing new ideas for further study of language and decision-making.