The importance of sustainable development has reached a consensus. Green consumption, as the final link of consumer behavior, can help green production activities make a real difference and achieve sustainable development. Based on the Agentic-Communal Model, this paper explores the relationship between power and green consumption through three experiments. The results showed that low-power (vs. high-power) consumers, who are more dependent on others, are likely to facilitate and encourage a communal orientation towards one's environment. These consumers pay more attention to others and may have a preference for green consumption. Self-concern plays a mediating role in this mechanism. However, when individuals have a strong impression management motivation, the difference in their willingness toward green consumption will disappear. In other words, both lower-power and high-power consumers are more willing to purchase green products. This paper helps to deepen the understanding of the psychological mechanisms underlying green consumption and also provides practical implications for firms' green marketing strategies.Sustainability 2019, 11, 4287 According to Baron and Kenny [71], in order to test the moderated mediating role of self-concern, 413 green consumption was regressed on the power and impression management motivation. The results 414 indicated that the interaction of power and impression management motivation was a significant 415 independent variable of green consumption with power, impression management motivation, and 416 power × impression management motivation as the independent variable (β = 0.18, p < 0.01). The 417interaction was also a significant independent variable of self-concern with power, impression 418 management motivation, and power × impression management motivation as the independent 419 variable (β = −0.09, p < 0.05). In addition, power × impression management motivation was also a 420 significant independent variable of green consumption with power, impression management 421 motivation, power × impression management motivation, and self-concern as the independent 422 variable (β = 0.15, p < 0.05). 423A bootstrapping analysis was used to test the moderated mediating role of self-concern. The 424 bootstrap mediating variable test was performed with reference to the mediation analysis model of 425 categorical variables proposed by Hayes (Model 8, n = 5000 samples) [72]. Under a 95% confidence 426 level, a consumer's self-concern was found to mediate the influence of power on green consumption. 427The results showed that the indirect effect (LLCI = 0. 0025 and ULCI = 0. 2411) does not contain 0. 428Furthermore, under weak impression management motivation, the mediating effect of a consumer's 429 self-concern was positive and significant (LLCI = −0.3649, ULCI = −0.1354). Under the strong 430 impression management motivation, the mediating effect of a consumer's self-concern was positive 431 and significant (LLCI = −0.2495, ULCI = −0.0457). 432 7. Conclusions and Discussion 433 7.1. Conc...