National surveys of teenagers have provided information for making decisions about smoking prevention programs. Repeated surveys are necessary because smoking patterns change over time. The prevalence of adolescent smoking and the variables related to it were investigated as well as the potential utility of those variables in predicting habitual smoking by adolescents. Two studies were conducted: the first was a 1979 survey of teenagers aged 12-18, their smoking behaviors, and related attitudes and demographics; the second was a 1979 follow-up study of a cohort sample first interviewed in 1974. Results indicated that the increasing prevalence of teenage smoking observed between 1968 and 1974 has subsided and a decrease in the smoking rates of both boys and girls has also occurred. Girls had a higher smoking rate than boys in 1979. The smoking practices of family members exerted a great deal of influence on the teenagers. Among their peers, teenage smokers associated with friends who smoked; nonsmokers also tended to associate with nonsmokers. (Author/HLM)