Chronic stress affects cognitive function across many domains, including memory, decision making and learning. While the effects of early-life stress on vocal learning in juveniles are well-demonstrated in both humans and songbirds, less is known about how stress experienced by adults affects their ability to learn new vocalizations or the neural substrates that underlie this behavior. We investigated the effects of chronic stress on the production and learning of contact calls, and on the expression of a key learning related gene, FoxP2, in the vocal learning circuit in adult budgerigars (Melopsittacus undulatus), a small parrot with open-ended vocal learning. We induced chronic stress via unpredictable disturbances in the captive environments of nine newly-formed replicate flocks of 4 adult male budgerigars who were previously unfamiliar to each other. We then recorded calling behavior daily and measured weight, breath rate, and baseline and stress response levels of circulating corticosterone weekly. At the end of the experiment brains were collected to examine mRNA and protein levels of the gene FoxP2 in the vocal learning region magnocellular nucleus of the medial striatum (MMSt) using qPCR and immunohistochemistry. Physiological measures of stress consistently showed stronger responses in birds subjected to the highest level of disturbance than those in the medium or baseline control treatments, although only differences in baseline corticosterone were detected among treatments. We used machine learning approaches to map calls onto a shared acoustic space to assess four measures of vocal behavior and learning, vocal output (the number of contact calls produced), vocal diversity (the amount of acoustic space occupied by the calls of an individual), vocal plasticity (the amount of change in acoustic space over time) and vocal convergence (the degree of overlap between an individual’s calls and the calls of its group). Birds in the high stress treatment showed higher vocal output and lower vocal plasticity than those in medium stress or baseline control groups, but there were no differences among treatments in vocal diversity or vocal convergence. There were no differences detected among treatments in expression levels of either FoxP2 mRNA or protein, perhaps due to the timing of neural sampling relative to the behavioral measures. These results suggest that, as seen in juvenile learning, chronic stress can negatively impact vocal learning in adults via changes in patterns of circulating corticosterone.