A naturalistic observation study of adolescent tobacco use was conducted to corroborate previous studies which used self-report questionnaire or structured interview methods to study this problem behavior. Several findings converged with previous accounts of adolescent tobacco use. For example, most use occurred in small groups. However, other findings diverged somewhat from previous research in that an unexpectedly low number of offers of tobacco was observed, nonusers were present in smoking groups, and solitary smoking was common. These data suggest that direct, normative social pressure to use tobacco may not be as frequent as more subtle informational social influence, at least in high school adolescent tobacco use groups.