PURPOSE To identify the determinants of PA of African-American schoolchildren ages 12–18 in Tunica County, Mississippi. A secondary purpose of this study was to determine the relationships among and identify differences between recognized determinants of PA (age, body mass index, perceived SE, and perceived barriers to PA) of the students. METHODS Assessment included height and weight measurements, completion of the Physical Activity Recall, self-efficacy and barriers to PA questionnaires. Subjects wore an MTI accelerometer for seven consecutive days, except when sleeping or bathing. Correlation matrices were conducted for male and female groups for the independent variables self-efficacy, barriers to PA, BMI and age, and the dependent variable, PA. RESULTS Nonsignificant correlations in the male group were found between PA and barriers to PA(r = −0.005, p = 0.977), and self-efficacy (r = 0.102, p = 0.0.527). Nonsignificant correlations in the female group were found between and between PA and barriers to PA (r = 0.411, p=0.119), and age (r = 0.101, p = 0.518). Stepwise multiple regression analyses were performed with independent variables that were related to the dependent variable, PA. In males (n = 41) the independent variables of age (r = 0.426, p < 0.05), and BMI (r = 0.543, p < 0.05), were regressed on the dependent variable, PA. The final regression model (with BMI and Age) accounted for 34% of the variance in PA. For females (n = 43), the independent variables of BMI (r = 0.508, p < 0.05), and self-efficacy (r = 0.411, p < 0.05), were regressed on the dependent variable, PA. It was discovered that the final regression model (with BMI and self-efficacy) accounted for 34% of the variance in PA. The Chow test was conducted to compare the regression equations for males and the regression equation for females. The comparison of both male and female regression equations' with the Chow test results in a nonsignificant regression model (F = 2.848, Fcv = 3.564). CONCLUSION The results of this study demonstrate that rural African-American adolescent males' PA is predicted by BMI and age but not self-efficacy and that females' PA is predicted by BMI and self-efficacy, but not age. The results also show that barriers to PA are not predictive of PA in rural African-American adolescents.