To help with the improvement of academic engagement, a structural model incorporating social and individual variables was deemed necessary. Hence, such a model was designed and evaluated with the help of a cluster sample of 511 high school juniors in Hamedan. A battery of questionnaires was used to collect data. The Anderson & Gerbing’s two- step approach was used to define the proposed structural model. This model well fitted the research population (X2/DF= 0.004, MSEA=1.87). Data analyses revealed that the effect of perceived social support on academic engagement is more indirect than direct, as it was assumed. Hence, developmental and psychological capitals, as well as psychological well-being are supportive mediators, although the role of psychological well-being as a second level variable, was not confirmed. All in all, the model explains 37% of the variance in academic engagement. It is thus concluded that increasing perceived social support can improve developmental and psychological capitals, and psychological well-being, leading to improved academic engagement.