Objective: To examine the influence of body mass index (BMI), sex, and race on college students' optimistic bias (OB) concerning the healthiness of their own lifestyles relative to the "average college student" and best friends. Design: Cross-sectional survey. Setting: Large university campus. Participants: College students recruited from first-year classes (1,398 female, 985 male). Variables Measured: Student height, weight, sex, race, and assessments of lifestyle healthfulness for self, best friends, and the "average college student." Analysis: Paired 2-way t tests compared lifestyle ratings for self and others. One-way analyses of variance compared differences between BMI categories. Tukey Honestly Significant Difference tests tested all pairwise comparisons. Analyses of covariance controlling for BMI scores tested sex and race differences (P less than 0.05). Results: College students in all BMI categories were optimistically biased in comparisons to the average college student. Optimistic bias differences were significantly smaller, however, for self and close friends. Sex and race also influenced the extent and nature of OB judgments. Conclusions and Implications: Optimistic bias may limit students' willingness to adopt and maintain more healthful behaviors. Optimistic bias can be reduced or even eliminated if close friends are specified as comparative targets. Health education campaigns, particularly those using socially normative approaches, should consider the impact of comparative target, sex, and race on OB. (Contains 4 tables.)