Birdsong is a useful model behavior for the study of the production of learned vocalizations. Its neural substrates are reasonably well mapped, but how the brain regions responsible for song generation accomplish this goal is still poorly understood. Much work has focused on the forebrain sensorimotor nucleus HVC, which is critical for song production and learning. Observations of extremely sparse bursts produced by projection neurons in HVC have been used to argue for a representation in this nucleus of time alone, with no relation to ongoing motor behavior.Other prior work in HVC has produced compelling evidence for rapid changes (in less than 4 hours) in intrinsic properties of a particular class of neuron in HVC as a result of a behavioral manipulation, delayed auditory feedback (DAF). As almost all examples of behavioral manipulation of auditory feedback in the zebra finch (a common model organism in birdsong research) require experimental durations of days to weeks, the possibility of a behavioral correlate of these intrinsic property changes was enticing. In this thesis, I examined both auditory feedback-related phenomena and HVC neuronal activity. First, I exposed birds to short periods of DAF and analyzed their singing behavior for changes, which were apparent nearly immediately. These changes were particularly related to the production of a preparatory behavior just prior to song onset. These results provide evidence of auditory feedback integration in zebra finch song at short latency, as well as for the use of auditory feedback to actively shape preparatory activity. Second, I reanalyzed a rich dataset of HVC neuronal activity and found that basal ganglia-projecting neurons in HVC produce bursts which comprise sequences during singing that appear to act as copies of one another, allowing information about neuronal activity in the past to be projected forward into the future. These recurring sequences could constitute a mechanism by which feedback, including auditory feedback, is integrated into ongoing HVC activity to produce a representation of song performance quality, a necessary precursor to correction of erroneous behavioral output. Third, I presented to sleeping birds auditory stimuli resembling delayed auditory feedback (pseudo-DAF), and observed responses involving both suppressed firing as well as potentiated activity relative to a well-studied auditory stimulus, the bird’s own song (BOS). These responses help to explain how DAF produces the changes in singing activity and neuronal intrinsic properties described above – abnormal neuronal responses interfere with ongoing HVC activity, producing error signals which ultimately result in behavioral change. Altogether, these results argue for the importance of auditory feedback to ongoing song production in the zebra finch. They also suggest likely features of neuronal responses to vocal production errors, and they outline a potential mechanism by which errors (assessed via auditory feedback or otherwise) are recognized and propagated within and beyond HVC.