Purpose: The aim of this study was threefold, to longitudinally examine the risk of habitual smoking onset in adolescents, to delineate the effects of neighborhood characteristics and friends’ smoking status on the habitual smoking onset, and to investigate whether the association between friends’ smoking status and habitual smoking onset was moderated by neighborhood characteristics. Methods: This study conducted multilevel discrete-time survival analysis, using cohort data from the 3rd to 6th waves of the Korean Child and Youth Panel Survey, which excluded habitual smokers, matched with 2010 census data on respondents’ residence. Results: Habitual smoking onset risk increased from the 8th to the 11th grade, and then slightly decreased from the 11th to the 12th grade. Friends’ smoking status (B = 0.60, p < .001), smoking rate (B = 0.06, p = .038), and the number of tobacco outlets in the respondents’ neighborhood (B = 0.51, p = .003) were positively associated with habitual smoking onset risk. Furthermore, the association between friends’ smoking status and habitual smoking onset risk was moderated by the number of tobacco outlets in the neighborhood. Specifically, the association was stronger in neighborhoods with more tobacco outlets (B = 0.58, p = .048). Conclusion: Friends’ smoking status and living in neighborhoods that are more susceptible to smoking increase the risk of habitual smoking. The number of tobacco outlets in the neighborhood enhances the peer effect of adolescent’s smoking behavior. Therefore, policies or interventions designed to reduce youth’s tobacco use should focus on not only on reducing peer smoking, but also restricting smoking by adults and the number of neighborhood tobacco outlets.