We examined the extent to which nicotine dependence and daily smoking might vary by age at first cigarette. The potential confounding effects of sex, race and history of childhood behaviour problems were examined as well. A sample of 1200 was randomly selected from the subset of 21-30-year-old members of a large HMO in the Detroit SMSA; 1007 (84%) agreed to participate. Personal interviews were conducted in respondents' homes, using the NIMH-DIS to elicit information on DSM-III-R diagnoses, including nicotine dependence. Controlling for sex and race, persons who smoked their first cigarette at 14 to 16 years of age were 1.6 times more likely to become dependent than those who initiated smoking at an older age (P = 0.03). The association was unchanged when history of childhood behaviour problems was also controlled. Smoking initiation before age 14 was not associated with increased probability of dependence. Persons who initiated smoking before age 14 had a longer lag time to daily smoking and a lower likelihood of progressing to daily smoking, compared to persons who initiated smoking later on. The findings suggest that, among persons who have ever smoked, there might be two distinct groups in whom the chances of developing dependence are considerably reduced. The first comprises persons who delayed first use until age 17. The second comprises persons who smoked their first cigarette before age 14, a group in whom the progression to daily smoking might be markedly slower than in persons who initiated smoking when they were older.