To examine the dimensions of unrealistic optimism, 635 college students indicated their relative likelihood of experiencing 25 health problems commonly used in studies of unrealistic optimism. Factor analyses of these estimates yielded five correlated, but distinct factors, suggesting that unrealistic optimism is not a unidimensional construct. Two factors, Common and Mixed, were comprised of heterogeneous problems. The Common factor was comprised of problems similar to those frequently used in other studies, and these problems were also perceived as more likely to occur than problems comprising other factors. Three factors were comprised of problems in specific domains, (1) substance abuse, (2) sexuality, and (3) mental health. Participants' psychological well-being and dispositional optimism were negatively correlated with perceived risk of experiencing mental health problems, whereas these measures were unrelated to perceived risks for other types of problems. The present results suggest that results of studies on unrealistic optimism may vary considerably as a function of the specific health risks being examined.Contemporary research on well-being is informed by the realization that wellness is influenced by the decisions and choices people make regarding health-relevant behaviors. Although some individuals behave in ways that promote well-being, others behave in ways that put them at risk for negative health outcomes. A variety of researchers have suggested that individuals engage in risky behaviors in part because they underestimate the likelihood that they will experience health problems, and such underestimates have been referred to as unrealistic optimism (Weinstein, 1980).In studies of unrealistic optimism, people are typically asked to estimate the likelihood they will experience a certain negative event relative 521