(Available in free full text) In this paper we use twin data from Australia to explore emotional well-being and its determinants. We aim to accomplish three things. First of all, using twin-fixed effects, and purging the estimates of common family environment and genetic similarities, we can test the robustness of previous findings in the well-being literature. We find that in the monozygotic twin-fixed effects estimations the marital status, health, years of education, and having low income preserve their significance, thus confirming the most pronounced stylized facts in the happiness literature. Second, using information about traumatic events, we test the validity of the adaptation hypothesis, according to which human beings can adapt to both positive and negative shocks and return to some setpoint level of life satisfaction. We find a strong negative effect of more recent traumatic events, such as being assaulted, being raped or being involved in an accident, which effects dissipate over time; thus, we confirm the validity of the adaptation hypothesis. Last but not least, we show that genetic dispositions are important for the within-pair variance of the emotional well-being.