started to decrease slightly (Monitoring the Future, 2002; Johnston et al., 2001). Youth in all sociodemographic groups experienced increased rates of smoking during the 1990s, although in the past several years, male adolescents have had higher rates of smoking than females, and white and Native American adolescents have had higher rates than other ethnic groups.

The process of becoming a regular or habitual smoker can be considered as a series of transitions through several stages, starting with a first “initiating” puff of a cigarette (Flay, Hu, and Richardson, 1998). A clear finding from research is that smoking initiation or experimenting is a behavior of youth. Among adult smokers, the vast majority tried their first cigarette as an adolescent, and most of these initiators proceeded to the next stages in the transition to habitual smoking before the age of 19 (Giovino, 1999). The most recent estimates (from 2001) suggest that, among eighth graders, 36.6 percent have ever tried a cigarette, 12.2 percent are current smokers, and 5.5 percent are daily smokers. Among twelfth graders, 61 percent have ever smoked, 29.5 percent are current smokers, and 19 percent are daily smokers. Despite the recent declining trend, current rates are similar to what was observed for youth nearly two decades ago in the early 1980s (Monitoring the Future, 2002).

The “social context” of adolescent tobacco use is of critical importance. Socioeconomic status is associated with youth smoking, with youth from families with lower levels of parental education and income more likely to smoke (Jacobson et al., 2001; Giovino, 1999). Also, factors such as the attitudes of parents and friends/peers toward smoking, whether or not friends or peers smoke, and whether or not a parent or other family member smokes are all significantly associated with youth smoking behavior (Jacobson et al., 2001; Richter and Richter, 2001). In addition, tobacco use does tend to cluster with other types of risk behaviors—including alcohol use—among adolescents (Windle and Windle, 1999; Bauman and Phongsavan, 1999). For example, youth who smoke are more likely to engage in binge drinking (Johnson, Boles, Vaughan, and Kleber, 2000).

Smoking rates among young adults (ages 19 to 25) also rose dramatically during the 1990s. Trend data from National Health Interview Surveys suggest that current smoking rates among young adults rose from 22.9 percent in 1991 to a peak of 27.9 percent in 1999 (Lantz, 2002). Several studies suggest that this increase was observed among both college students and young adults not in school (Wechsler, Rigottu, Gledhill-Hoyt, and Lee, 1998; Johnston, O’Malley, and Bachman, 2001; Lantz, 2002). Although smoking rates among young adults also have declined recently, the dramatic increase observed during the 1990s remains alarming. Based on timeseries analysis (Lantz, 2002), it does appear that approximately 75 percent of this increase is due to a cohort effect: that is, the observed increase in smoking among young adults reflects in large part the aging of adolescent