This study examined whether and at what time point various interpersonal and intrapersonal correlates of smoking shifted over four assessment points in a longitudinal study of 461 individuals. The study began when the subjects, who had been participating in a study of adolescent and young adult development were in junior high school. Four assessment points, 4 years apart were available for this study. Latent variables representing Cigarette Use, Social Relations, Extraversion, and Depression were intercorrelated within each time period. As hypothesized, a substantial shift in the patterns of intercorrelations was found over time. At Time 1, when the subjects were in junior high school and may be described as experimenters, Cigarette Use was positively correlated with good Social Relations and Extraversion, and negatively with Depression. Only four years later Cigarette Use was significantly and positively correlated with Depression, negatively correlated with good Social Relations, and only minimally correlated with Extraversion. At Times 3 and 4 when subjects were in their early and late 20s the same pattern emerged of a significant positive correlation between smoking and Depression, a negative correlation with good Social Relations, and a minimal correlation with Extraversion. These results indicate that smoking in response to peer pressure alone may not continue; those who continue smoking may use it to relieve tension and for self-medication. (Author/ABL)