The cochlear nucleus is the gateway for central nervous system processing of auditory information in mammals. It has been proposed that parallel processing channels are set up in the CN, and these form the basis for further computation at higher stations of the auditory system. Despite decades of study, enumeration of CN cell types is incomplete and CN circuitry is described only superficially. In neuroscience generally, classification and naming of neurons has relied primarily upon qualitative approaches based upon human observational capabilities. We have implemented and in some cases developed novel high-throughput and unbiased techniques for labeling, segmenting and classifying neurons in 3D, generated from large-scale electron microscopy image volumes. We propose to deliver a nanoscale map, or connectome, of the mouse CN with enumerated and localized cell types and their synaptic connections. This effort is unbiased because all neurons will be sampled. To achieve this goal, we bring together four parallel modes of tissue analysis for neuron classification: morphology, connectivity, molecular identity and function. We propose that connectivity analysis will define long-proposed parallel processing circuits that will be tested functionally using realistic biophysical models of identified cell types. Notably, the cochlear nucleus contains both amorphous and layered organizations of cells, which serve as templates for all other brain regions. By investigating the fundamental structure of this sensory center, we will establish principles of neural computation and methods for structural and functional phenotyping that will apply to other brain regions regardless of their particular neural architecture.