The theory of risk compensation predicts that when individuals increase the practice of a safe behavior, they will also increase the practice of an unsafe behavior in order to maintain the same level of overall risk. In contrast, response generalization predicts that an increase in safe behavior will increase other safe behaviors in the same response class. The present study examined the effects of community-based feedback on the safety belt use of drivers in a small community in Southwest Virginia, while drivers on an intersecting highway served as a control group. An AB design was used to test the effects of the feedback on belt use. Turn signal use and right and left hand turn behaviors were also measured to study risk compensation vs. response generalization. Baseline measures were taken for 13 weeks at which point two feedback signs were erected in the community for the remaining 17 weeks of the study. The words "SAFETY BELT USE IN NEWPORT LAST WEEK" with the percentages of male and female driver safety belt use the previous week appeared on the signs. Feedback increased the safety belt use average in the community by 15.5 percentage points for males, and 9.7 percentage points for females over a 17 week period. Evidence for response generalization was shown by a 14.9 percentage point increase in turn signal use over the 17 weeks of the feedback intervention.