American Indians (AI) have high prevalence of diabetes, which might be reduced through targeted lifestyle changes. We conducted an exercise program for AI adolescents at risk for diabetes. We tested whether the participants’ perceptions of physical activity were related to their physical characteristics or exercise participation. Overweight (OW) AI youth (n = 75), 11-20 years old, with low physical activity, were instructed to exercise 3d/week for 16 weeks at tribal centers. Another 39 normal weight (NW) peers from the same communities served as a reference group but did not enter the exercise program. Perceived comfort and enjoyment of physical activity was measured on a 15-item, 5-point scale. Clinical tests included body composition, aerobic fitness (VO2pk), and physical activity (daily steps). Mean perception score was higher (more positive outlook) for NW than OW/Ob (4.1 ± 0.5 vs. 3.5 ± 0.6, p < 0.001) and for boys than girls (3.9 ± 0.5 vs. 3.5 ± 0.6, p < 0.02). Perception was negatively correlated with body mass index (r = -0.52), percent body fat and total fat free mass (r = -0.54 and -0.23, respectively), and positively correlated with VO2pk and daily steps (r = 0.53 and 0.38). OW participants completed 27 ± 15 exercise sessions, lasting 34 ± 7 minutes per session. Participants who completed at least 24 sessions had higher perception scores than those with less than 24 sessions (3.6 ± 0.6 vs. 3.3 ± 0.6, p=0.04). More OW participants reported perceived limitations to exercise (64%) and life stressors (85%) than the NW group (51% and 74%, respectively), but these barriers did not explain variance in exercise compliance. In AI adolescents, perception of physical activity varies with body weight, physical activity, and aerobic fitness, and may contribute to rates of participation in a wellness-center based exercise program. Thus, physical activity programs to reduce diabetes risk in this population must be tailored to address additional barriers related to one’s perceptions in order to to optimize participation. Disclosure H. Kimbley: None. J. Chadwick: None. C.L. Coleman: None. M.A. Tullier: None. L.D. Wolbert: None. K.C. Copeland: Other Relationship; Self; Novo Nordisk Inc.. K.R. Short: None.