Research on the consequences of optimism has tended to look at optimism as a stable dispositional characteristic, defining individuals along a single continuum and typically neglecting the possibility that people's predictions may vary across situations. However, other lines of research have shown that optimism can and does vary from situation to situation, suggesting that there may be more to optimism than always expecting the best. Using a new measure of optimism, the primary goals of this thesis are (1) to determine whether optimism can be understood not only in terms of how optimistic people tend to be in general (overall level), but also in terms of how much optimism tends to vary from situation to situation (cross-situational flexibility); and (2) to evaluate both dimensions of optimism by comparing people's descriptions of their own predictive tendencies to conventional notions of what ideal predictions ought to be (correspondence to prescribed ideals). I hypothesize that both overall level and cross-situational flexibility are measurable dimensions of optimism, as is correspondence to a situation-specific prescribed ideal, and that all three individual difference variables represent reliable aspects of a person's orientation toward the future. I also hypothesize that not only does the expression and experience of optimism vary by situation, but that the ideal prediction also varies, and corresponding to this shifting ideal is important for well-being. The "wise optimist" is predicted to be generally optimistic, but not too optimistic, and flexible enough to adjust their responses appropriately depending on the specific demands a given situation. To test these hypotheses, I conducted two studies: A cross-sectional study and a longitudinal study utilizing a cross-lagged panel design. The results of the cross-sectional study (N = 347) revealed that all three variables (level, flexibility, and correspondence) are measurable and reliable, and when entered into a simultaneous multiple regression, each was found to make a significant, independent contribution to well-being. The longitudinal study (N = 233), which utilized a more comprehensive measure of well-being, allowed me to investigate questions of both causality and stability (test-retest reliability) over time. Using structural equation modeling to evaluate the longitudinal data, correspondence to prescribed ideals was found to be a significant predictor of well-being over time, suggesting that there is some real wisdom to conventional wisdom, and corresponding to ideals prescribed by others has some important implications for well-being, at least in terms of subjective well-being. Taken together, these studies suggest that there is much more to being a wise optimist than always expecting the best.