The data show that being of indigenous or mixed ethnicity constituted significant risk factors in the initial and middle stages of the tobacco smoking trajectory, even after controlling for potential confounders. Compared with youth of European background in the Jujuy sample, youth of indigenous and mixed ethnicity had a higher prevalence of current smoking (23 percent for both indigenous and mixed ethnicity compared to 11.4 percent for youth of European background); however, ethnicity was not significant in the more advanced stage of smoking, i.e., established smoking, where other factors may play a more important role, such as the addictive power of nicotine, alcohol consumption, or mental health problems such as depression. There was no significant difference in current smoking rates for boys and girls within the ethnic groups studied. In addition, there was no evidence of an association between smoking and parental education level or employment status, except for a lower risk of ever smoking among those with unemployed parents. Psychosocial risk factors consistently associated with smoking were similar to those reported in the literature, with the most influential being peer pressure, thrill-seeking orientation, and mental health problems. The study involved a self-administered survey of 3,131 eighth-grade youth enrolled in a random sample of 27 urban and rural schools in 2004 in Jujuy, Argentina. Standard questions adapted from global surveys were used. 2 tables, 1 figure, and 44 references