Over the last decade, there has been an explosion in cigarette smoking rates among college students. Smoking in adolescent populations has been associated with a number of environmental, personal, and behavioral factors, but smoking in the college-age population remains virtually unexplored. The purpose of this study was to determine the factors that predict smoking group membership in a college-age population. A non-experimental cross-sectional design was used to examine factors that distinguished among four groups: never established (n = 591), former (n = 41), non-daily ( n = 80), and daily (n = 76) cigarette smokers. The sample consisted of 788 randomly selected college students, 18 to 24 years of age, who completed and returned a pencil and paper questionnaire. Significant differences were found among the groups on their mothers', fathers', and siblings' and peers' smoking behaviors and attitudes toward smoking. Daily smokers were the most likely group to report their mothers smoked. Former smokers were least likely to report their fathers smoking, and the never established group reported the lowest exposure to smoking siblings. Parents' attitudes toward smoking did not differ among the four groups. The two current smoking groups reported a greater number of risk behaviors than the two non-smoking groups. Significant differences also were found in the age and number of depressive symptoms reported by the groups. Older students were more likely to report daily or former smoking, and the two current smoking groups were more likely to report an increased level of depressive symptoms. The number of high school and college friends who smoke, alcohol and other drug use predicted smoking group membership using a polytomous logistic regression. Strategies to alter the social environment and decrease the exposure of non-smoking students to others who model smoking behaviors may help decrease smoking initiation and prevent relapse on college campuses.