AbstractNeural activity within the cortical premotor nucleus HVC (acronym is name) encodes the learned songs of adult male zebra finches (Taeniopygia guttata). HVC activity is driven and/or modulated by a group of five afferent nuclei (the Medial Magnocellular nucleus of the Anterior Nidopallium, MMAN; Nucleus Interface, NIf; nucleus Avalanche, Av; the Robust nucleus of the Arcopallium, RA; the Uvaeform nucleus, Uva). While earlier evidence suggested that HVC receives a uniformly distributed and nontopographic pattern of afferent input, recent evidence suggests this view is incorrect (Basista et al., ). Here, we used a double‐labeling strategy (varying both the distance between and the axial orientation of dual tracer injections into HVC) to reveal a massively parallel and in some cases topographic pattern of afferent input. Afferent neurons target only one rostral or caudal location within medial or lateral HVC, and each HVC location receives convergent input from each afferent nucleus in parallel. Quantifying the distributions of single‐labeled cells revealed an orthogonal topography in the organization of afferent input from MMAN and NIf, two cortical nuclei necessary for song learning. MMAN input is organized across the lateral‐medial axis whereas NIf input is organized across the rostral‐caudal axis. To the extent that HVC activity is influenced by afferent input during the learning, perception, or production of song, functional models of HVC activity may need revision to account for the parallel input architecture of HVC, along with the orthogonal input topography of MMAN and NIf.