Problem/Objective: Tobacco industry market research identifies targets for tobacco marketing using “psychographic” characteristics. We adapted research questionnaires from previously secret industry documents in a segmentation study of young adults (age 18-25). Methods: Cross sectional national panel survey of 1528 young adults using a panel derived from a Random Digit Dialing sample (response rate 79.3%). Questions included attitudes about the future, personality characteristics, self-descriptors, activities, and smoking attitudes. Factor analysis followed by cluster analysis was used to describe naturally occurring segments of the population. Results: We found two clusters with high smoking rates (42% and 57% current (past 30 days) smokers) and two with low smoking rates (10% and 13%). One low risk cluster was risk averse, had traditional roles, and was not receptive to tobacco advertising. Another low risk cluster had an independent, mature self image, and strongly believed secondhand smoke is dangerous. One high risk cluster was very receptive to tobacco advertising, felt smoking was useful, and disbelieved dangers of secondhand smoke. The other high risk cluster was highly social, frequently participated in social and physical activities, and valued being cool, popular, and unique. Both high risk clusters reported greater alcohol use. High risk cluster membership was significantly associated with current and established smoking in multivariate logistic regressions controlling for demographic variables. Conclusions: Tobacco market research surveys can identify high risk subgroups of the young adult population suitable for targeted tobacco control programs. Different high risk groups may require different message strategies to decrease their tobacco use.