Objectives. Using a sample of elementary and middle school students, we examined the associations between body mass index (BMI), obesity, and measures of the proximity of fast food and full service restaurants to students' residences. We controlled for socioeconomic status using a novel proxy measure based on housing values. Methods. We used BMI and obesity measures based on height and weight data collected as part of a school health assessment along with geocoded data on addresses of residences and food establishments. We constructed a proxy measure of socioeconomic status from public records of residential property assessments. These data were used to estimate logistic regression models of overweight and ordinary least squares models of BMI. Results. Students residing in homes with higher assessment values were significantly less likely to be obese, and had significantly lower BMIs. Upon controlling for socioeconomic status and other characteristics, the associations of BMI and obesity with proximity to food service establishments were reduced. Nonetheless, students who resided within one-tenth or one-quarter of a mile from a fast food restaurant had significantly higher values of BMI. The proximity of full service restaurants to residences did not have a significant positive association with either BMI or overweight. Conclusion. Public health efforts to limit access to fast food among nearby residents could have beneficial effects on child obesity. Public data on property value assessments may serve as useful approximations for socioeconomic status when address data are available.