Converging evidence implicates the auditory forebrain regions caudal medial mesopallium (formerly cmHV) and caudal medial nidopallium in the perceptual processing of conspecific vocalizations in songbirds. Little is known however, about more specific processing within these regions especially during song-based perceptual behaviors. One hallmark of the caudal medial mesopallium and caudal medial nidopallium, areas analogous to mammalian secondary auditory cortical structures, is their robust expression of the immediate-early-gene zenk in response to conspecific songs. Using European starlings operantly trained to recognize the songs of individual conspecifics, we show that the levels and patterns of zenk protein expression in the caudal medial nidopallium and caudal medial mesopallium differ when song recognition demands are placed on the system. In the caudal medial mesopallium, expression is significantly elevated above basal levels during the recognition of familiar songs, the acquisition of novel associations for familiar songs, and the acquisition of novel song discriminations. In the caudal medial nidopallium, however, expression is significantly elevated above basal levels only during the acquisition of novel song discriminations. The results directly implicate the caudal medial nidopallium and caudal medial mesopallium in at least a portion of the auditory processes underlying vocal recognition. Moreover, the observed differences between these regions imply the functional localization (or at least the concentration) of different auditory processing mechanisms within the caudal medial nidopallium and the caudal medial mesopallium.