Existing research showed that altruistic behavior was conducive to the improvement of subjective well-being. Internet altruistic behavior refers to an act performed voluntarily on line to help someone else when there is no expectation of receiving a reward of any form. However, few explored the relationship between internet altruistic behavior and subjective well-being, which was beneficial to academic development, the improvement of life quality and physical and mental health among adolescents. Thus, the first aim of the study was to examine the relationship between internet altruistic behavior and subjective well-being. Moreover, the formation of subjective well-being was influenced by many factors. As a relatively stable personality, self-esteem was one of the most reliable and the most powerful predictors of subjective well-being. According to the theory of sociometer, self-esteem could adjust the process of individual evaluating. Therefore, we speculated that self-esteem could prompt individual to form a higher self-assessment after making internet altruistic behavior, so we tested whether self-esteem served as a moderator of the relationship between internet altruistic behavior and subjective well-being. Considering that people with high self-esteem had higher self-efficacy compared to that with low self-esteem when it comes to success or failures, we also tested the moderating effect of self-esteem may be completely or partially mediated by self-efficacy. A sample of 1376 adolescents (656 males, 720 females) with mean age 14.83±1.69 years, mean online time a week 5.48±3.21 hours was randomly recruited for the study. Participants were tested with Internet Altruistic Behavior Scale, General Self-Efficacy Scale, Self-Esteem Scale and Subjective Well-Being Scale. The present study examined the associations between internet altruistic behavior, general self-efficacy, self-esteem and subjective well-being. Meanwhile, hierarchical regression analyses were applied to detect the effects of internet altruistic behavior on subjective well-being as well as the moderating role of self-esteem and the mediating role of self-efficacy. Our findings are as follows: (1) Significant common method bias were not exist in the study; (2) There were significant positive correlation between internet altruistic behavior and subjective well-being, which ensured that the supposed variables could be analyzed as mediators; (3) Regression analysis showed that self-esteem acted as a moderator between the relationship of internet altruistic behavior and subjective well-being. That is, subjects with high self-esteem showed significantly higher subjective well-being levels relative to subjects with low self-esteem, and the ability of internet altruistic behavior to predict subjective well-being was weakened for subjects with low self-esteem. Additionally, the moderating effect of self-esteem was completely mediated by self-efficacy. In summary, this study contributes to the literature in several ways. On one hand, the integration of the mediating and moderating models provides a more comprehensive and elaborative interpretation to understand how and when internet altruistic behavior impacts subjective well-being. On the other hand, our findings provide a theoretical and empirical basis for others to study the complicated relationship between internet altruistic behavior and subjective well-being. Last but not the least, this study is helpful for future work on measures to improve adolescents’ subjective well-being. More attentions should be paid to the low internet altruistic behavior individuals and the low self-esteem individuals.