This paper focuses on the cognitive aspect of self-consciousness, self-reflection, and on subjective well-being. Based on empirical evidence that a linear relationship between these variables is questionable, alternative pathways are described and tested. First, is there a curvilinear relation (Study 1)? Second, does self-reflection become a risk factor for well-being when it is used as a rigid metacognitive strategy (Study 1)? And third, does self-reflection interact with self-efficacy (or other variables) to predict well-being (Study 2)? Finally, possible relationships between the variables are briefly reviewed from an interactionist perspective.