Features of tobacco dependence vary by ethnicity, which could be partially due to measurement bias inherent in instruments that assess nicotine dependence. This study compared responses on the Fagerström Test for Nicotine Dependence (FTND) by African American adolescents (n = 478) with those of White adolescents (n = 661) seeking smoking cessation treatment. We conducted item-by-item comparisons by ethnicity for the six questions composing the FTND and confirmatory factor analyses and multiple indicators-multiple causes (MIMIC) modeling to test the hypothesis of measurement invariance of the FTND by ethnicity. Study participants (N = 1,139) were daily smokers of average age 15.4 years (SD = 1.3); 42.0% were African American and 61.5% female. White adolescents' pattern of responses to the FTND indicated a greater degree of dependence; these differences were statistically significant for five of the six items. The FTND exhibited a unidimensional structure with similar factor loadings in both White and African American adolescents. However, MIMIC modeling indicated differential reporting for three out of six items, suggesting that the FTND may not measure nicotine dependence equivalently for White and African American youth.