Recent reports have indicated a substantial increase in adolescent obesity. This increased prevalence is of significant concern given evidence that youth obesity tends to track into adulthood and that adults who were obese as children experience increased morbidity and mortality independent of adult weight. The purpose of this study was to determine which variable tested, either alone or in combination; best predicts obesity in high school students of different races. The study population consisted of 1628 female students drawn from schools in a large Midwestern city and suburb. Students were given the Fitnessgram® fitness test and completed a 15 question survey instrument designed to assess barriers to exercise and activity level. 1314 students (White, N=641; African American, N=367; Hispanic, N=306) completed all aspects of the study and were used in the final analysis. Hierarchical Regression Analyses were conducted on each race to determine the amount of variance that the variables: barriers to exercise, mile run time, and activity level contributed to obesity as measured by percent fat. For African American students, the variables combined accounted for 58% of the variance (barriers to exercise accounted for 50% while the mile run added 5% and activity level contributed an additional 3%). For Hispanic students, barriers to exercise accounted for 14% of the variance while mile run time and activity level contributed 10% and 6% respectively. For White students, the mile run accounted for 53% of the variance and activity levels an additional 17%. Health promotion programs that target potential barriers to physical activity, and increases in students' actual physical fitness levels may help to reduce the increasing prevalence of adolescent obesity.