Objective: To identify where and how middle school students acquire cigarettes and to describe factors associated with a decreased likelihood of being asked for identification during the purchase of cigarettes. Methods: A modified version of the CDC's Youth Risk Behavior Survey was adminstered to 2,227 6th-8th grade students attending 53 randomly selected middle schools in NC. Data were analyzed with Cramer's V, Kruskal-Wallis ANOVA, and multiple logistic regression using the LR approach. Results: 497 (22.8%) of the students had smoked during the previous 30 days and this subsample was the focus of the analyses. 58% of current smokers were male and 69% were white. Only 16.9% of the students had purchased cigarettes from a store and 3.9% had purchased cigarettes from a vending maching during the previous 30 days. Females were more likely (p<.006) than males to borrow or to have someone else buy cigarettes for them. Males were more likely (p<.006) than females to buy cigarettes from a store or vending machine or to steal them. Students who purchased cigarettes from a store or vending machine were more likely(p<.032) than other smokers to be one or more years older than expected for their school grade. Among the 197 students who had purchased cigarettes during the previous 30 days, only 14.2% had beed asked for proof of age. Students who acquired cigarettes from a store were more likely to have not been asked for proof of age, while having been asked for proof of age was associated with acquiring cigarettes by other means (V=.44, p<.0001). Not being asked for proof of age was associated with earlier age of onset of smoking (V=.34, p<.0001), number of days during the previous month cigarettes were smoked(V=.34, p<.0001), and number of cigarettes smoked per day (V=.35, p<.0001). When analyzed with multiple logistic regression, only earlier age of onset of smoking remained a significant correlate of not being asked for proof of age. As age of onset of smoking increased, the odds of being asked for proof of age increased 1.36 (95% CI= 1.16-1.60) times per year of age, with the reference year being 9 years of age or younger. Conclusion: In NC it was very easy for these young adolescents to illegally purchase cigarettes, especially for more experienced smokers. However, being asked for proof of age forced these youth to acquire cigarettes using other methods, and this was associated with smoking less. Prevention efforts should include enforcing existing laws requiring youth to provide proof of age when buying cigarettes.