Abstract Background Tobacco use is a leading cause of preventable morbidity and mortality. In the developed nations where the burden from infectious diseases is lower, the burden of disease from tobacco use is especially magnified. Understanding the factors that may be associated with adolescent cigarette smoking may aid in the design of prevention programs. Methods A secondary analysis of the 2004 United States National Youth Tobacco Survey was carried out to estimate the association between current cigarette smoking and selected smoking-related variables. Study participants were recruited from middle and high schools in the United States. Logistic regression analysis using SUDAAN software was conducted to estimate the association between smoking and the following explanatory variables: age, sex, race-ethnicity, peer smoking, living in the same household as a smoker, amount of pocket money at the disposal of the adolescents, and perception that smoking is not harmful to health. Results Of the 27727 respondents whose data were analysed, 15.9% males and 15.3% females reported being current cigarette smokers. In multivariate analysis, compared to Whites, respondents from almost all ethnic groups were less likely to report current cigarette smoking: Blacks (OR = 0.52; 95% CI [0.44, 0.60]), Asians (OR = 0.45; 95% CI [0.35, 0.58]), Hispanic (OR = 0.81; 95% CI [0.71, 0.92]), and Hawaii/Pacific Islanders (OR = 0.69; 95% CI [0.52, 0.93]). American Indians were equally likely to be current smokers as whites, OR = 0.98 [95% CI; 0.79, 1.22]. Participants who reported living with a smoker were more than twice as likely to smoke as those who did not live with a cigarette smoker (OR = 2.73; 95% CI [2.21, 3.04]). Having friends who smoked was positively associated with smoking (OR = 2.27; 95% CI [1.91, 2.71] for one friend who smoked, and OR = 2.71; 95% CI [2.21, 3.33] for two or more friends who smoked). Subjects who perceived that it was safe to smoke for one or two years were more likely to smoke than those who thought it was definitely not safe to do so. There was a dose-response relationship between age and the amount of money available to the respondents on one hand, and current smoking status on the other (p-value < 0.001). Conclusion We found that White non-Hispanic adolescents were as likely to be current smokers as American Indians but more likely to be smokers than all other racial/ethnic groups. Older adolescents, increase amounts of pocket money, and perception that smoking was not harmful to health. The racial/ethnic differences in prevalence of smoking among America youth deserve particular exploration.