either overestimate (Borland 1997; Kristiansen et al. 1983; Viscusi 1990; 1991; Viscusi et al. 2000) or underestimate (Schoenbrun 1997; Sutton 1997) the particular risks of smoking compared with the actual risk from epidemiological data, most studies agree that adolescents and young adults are aware of many of the risks involved with tobacco use. In particular, they are aware that smoking involves a significant risk of lung cancer and other health outcomes (Jamieson and Romer 2001a). However, the literature also indicates that adolescents are not aware of the full extent to which smoking is harmful (Arnett 2000; Covington and Omelich 1992; Eiser and Harding 1983; Halpern-Felsher et al. 2004; Hansen and Malotte 1986; Leventhal et al. 1987; Virgili et al. 1991), nor do they fully understand the extent to which tobacco use can shorten the life span (Romer and Jamieson 2001a). More importantly, adolescents are less likely to believe that the risk of addiction and the related health consequences apply to them. To complicate matters, adolescents show an incomplete understanding of the addictive nature of tobacco use that is related, in part, to their inaccurate assessment of smoking risks and their belief that they can quit at any time and therefore avoid addiction (Arnett 2000; Slovic 1998). Furthermore, they believe that smoking risks can be counteracted by altering the amount that they smoke, when they smoke, or what they smoke (e.g., “light” versus regular cigarettes) (Kropp and Halpern-Felsher 2004).

Whether such perceptions, or misperceptions, actually motivate or predict tobacco use is a complicated question. For example, although the concept of adolescent invulnerability is widely used to explain why adolescents smoke, the few studies that have examined the relationship between personal risk perceptions and tobacco use have yielded mixed results. Although some studies find that adolescents who have smoked perceive greater personal risks (Gerrard et al. 1996; Johnson et al. 2002; Resnicow et al. 1999), others show that smokers perceive less personal risk (Arnett 2000; Covington and Omelich 1992; Eiser and Harding 1983; Goldberg et al. 2002; Urberg and Robbins 1981; 1984; Virgili et al. 1991). The following sections details smokers’ (particularly adolescent smokers’) risk-related beliefs regarding several such aspects of tobacco use.

Adolescent smokers tend to overestimate some smoking risks and underestimate others. In general, they understand that smoking causes lung cancer, but they also overestimate the degree to which it does. Jamieson and Romer reported that among 14- to 22-year olds surveyed in the Annenberg Tobacco Survey, 70 percent of smokers and 79 percent of nonsmokers overestimated the risk of lung cancer attributed to smoking (Jamieson and Romer 2001b). The survey respondents also underestimated the degree