DESCRIPTION (provided by applicant): The proposed study will build on information collected in our previous study, "Community Characteristics and Physical Activity among Adolescent Girls", which was an ancillary study to TAAG, Trial of Activity of Adolescent Girls. We will follow the same girls from the Minneapolis and San Diego TAAG sites and identify where they go when they are not in school or home by using portable global positioning system (GPS) units, unobtrusive devices worn on the wrist or hip that record the latitude and longitude of locations at user-specified time intervals. GPS will be combined with accelerometry and self-reports of food purchases and consumption so that we will be able to determine the context in which physical activity and eating occurs in neighborhoods. In the baseline study we found associations between neighborhood characteristics such as parks and playgrounds and physical activity, but could not determine if girls were active in these settings or whether increased levels of physical activity were due to community norms. Our specific aims are: 1) To identify the specific environmental exposures to food sources and activity venues based upon actual travel patterns of older adolescent girls. 2) To study the trajectory of physical activity in older adolescent girls, how physical activity patterns change when girls and their peers become more autonomous, and to identify the environments in which girls are physically active and the characteristics of those environments, 3) To study the trajectory of BMI among adolescent girls as they grow into adulthood and determine if eating away from home is associated with neighborhood food environments, BMI, and with changes in BMI. We will obtain both dietary and physical activity data from 300 girls at two timepoints, half will be in 10th grade at the first measurement and the other half in 11th grade. The second measurement will be in the 11th and 12th grades in order to understand how diet and physical activity change as girls become more autonomous, obtain driving licenses and spend more time in cars alone or with peers. We will analyze the data using propensity scoring and longitudinal growth models to identify how local environments may play a role in physical activity, diet and BMI. The findings will be useful to inform future community and population level interventions to promote healthy eating, active living, and obesity control.