People tend to maintain certain positive illusions about themselves and their futures that may be adaptive in buffering their self-esteem and feelings of efficacy from the effects of negative or threatening feedback. The illusion of unique invulnerability is the expectation that others will be the victims of misfortune and negative events more so than oneself. One possible implication of holding this belief is that, if a false sense of security is fostered, actual vulnerability to experiencing negative events that one has control over (for example, contracting lung cancer from smoking) might be increased if self-protective behaviors are decreased. The purpose of this study was to determine 1) what personality characteristics are related to this belief in unique invulnerability, 2) whether unique invulnerability is related to assumption of risk in behavior, 3) how accurately people assess the riskiness of their own behavior, and 4) how beliefs in unique invulnerability change over time and experience with risk. Subjects were 164 male and female Reserve Officer Training Corps (ROTC) cadets, ages 17-20, engaged in rappelling down a tall structure, an inherently risky activity. A measure of unique invulnerability and several personality measures were administered before rappelling, and risk-relevant behavior was observed during rappelling. Personal safety ratings were obtained and the invulnerability measure was readministered after rappelling. Results indicated that unique invulnerability was positively correlated with self-esteem (p<.004). For subjects’ first rappels, unique invulnerability was negatively related to number of bounds to descend a 40-foot tower (p<.01), such that as invulnerability increased, number of bounds decreased. Although assumption of risk was operationally defined as taking fewer bounds, further analyses indicated that number of bounds might be more a reflection of physical competence than of riskiness. This finding suggested that greater beliefs in invulnerability were related to competent rappelling performance, while lesser beliefs in invulnerability were related to ultra-conservative behavior. Self-reports of safety were positively related to number of bounds for first rappels off a 40-foot tower (p<.01), and were related to instructor ratings of jump competence and safety for first and second rappels off a 40-foot tower (ps<.05). This suggests that as a number of bounds became more conservative, and as instructor ratings improved, self-reports of safety increased. Finally, beliefs in unique invulnerability decreased after rappelling, both for subjects who rappelled (p<.001), and for control subjects who observed but did not rappel (p<.05). Conclusions are that self-esteem is an important component of the illusion of unique invulnerability, that invulnerability is related to competent performance while engaging in risky activities like rappelling, that invulnerability is subject to change over time and experience, and that subjects can fairly accurately assess the objective safety of their behavior. Implications for theory and research on unique invulnerability are discussed, as well as limitations and future directions. / Master of Science