There is a growing body of evidence indicating that urban sprawl, poor land use-mix, lack of infrastructure conducive to physical activity (PA), and negative perceptions about neighborhood safety and aesthetics are related to reductions in PA and to subsequent increases in obesity. With approximately one in three university students classified as overweight or obese, PA and the prevention of obesity are the top two priority health indicators in the Healthy Campus 2010 initiative. Few studies, however, have examined relationships between the physical environment and PA behavior in university students, and the link with subsequent health outcomes has not been investigated. Moreover, in built environment-obesity studies only self-reported body mass index (BMI) values have been used as the criterion measure for obesity risk. While BMI values provide an indication of body fatness, methods such as total body dual-energy X-ray absorptiometry (DXA) provide a more accurate estimation of adiposity. The purpose of this study was to examine the associations between perceptions of the physical environment and percent body fat (%BF) estimated from DXA. Participants were undergraduate student volunteers enrolled in a southern university. The Neighborhood Environment Walkability Scale (Saelens, et al., 2003) was used to assess perceptions of the physical environment. Percent body fat was estimated from total body DXA scans (Prodigy Pro, GE Medical Systems, Madison, WI). Stepwise regression analyses were conducted to examine relationships between eight specific aspects of the physical environment (residential density, land use mix-diversity, land use mix-access, street connectivity, walking/cycling facilities, aesthetics, traffic safety and crime safety) and %BF. Sixty-one participants completed testing (M age = 18.59 years, SD = 0.56). Twenty-eight were male (25 White; 2 Black; 1 Hispanic) with a mean %BF of 17.75 ± 8.63. Thirty-three participants were female (23 White; 10 Black) with a mean %BF of 32.33 ± 5.98. For males only, land use mix-access and crime safety accounted for 33% of the variance in %BF (F2,25 = 6.27, p = .01). In this group of male college students, those who perceived their neighborhoods as having limited access to destinations (bus stops, stores, etc.) and high crime had a higher %BF. The relationships evident in this age group using a more precise method of estimating body fatness support previous research which used BMI as the criterion measure. These results suggest that in a college-age population, aspects of the neighborhood environment are associated with risk for obesity and gender-specific differences may exist. Keyword(s): college level issues, health promotion, obesity issues