Traffic noise is a prevalent and yet poorly understood anthropogenic disturbance associated with reduced avian diversity, population densities and pairing and mating success. How these systems are affected is not clear as a direct experimental link between noise and behavior underlying these patterns is missing. Here we provide the first empirical evidence of the effects of long-term exposure to simulated traffic noise on social and risk-taking behavior of Carolina chickadees (Poecile carolinensis) and tufted titmice (Baeolophus bicolor). In testing for these effects we compare two hypotheses regarding the effects of noise on behavior. We found that noise increases sociality by reducing nearest neighbor distances and increasing the number of close-perches within study flocks. These behavioral responses mimic those of species in high-risk situations, such as birds in the presence of a predator. These results provide support for the 'Increased Threat Hypothesis,' which argues that chronic traffic noise affects behavior by increasing the perceived level of threat. Although the adaptive value or function of these responses to noise is unknown, they may serve to mitigate any negative effects of traffic noise. If true, species lacking behavioral plasticity may be more susceptible to effects of traffic noise and other similar acoustic disturbances.