The suprachiasmatic nucleus, site of the dominant mammalian circadian clock, contains a variety of different neurons that tend to form groups within the nucleus. The present investigation used single and multiple label tract tracing and immunofluorescence methods to evaluate the relative locations of the neuron groups and to compare them with the distributions of the three major afferent projections, the retinohypothalamic tract, geniculohypothalamic tract and the serotonergic pathway from the median raphe nucleus. The suprachiasmatic nucleus has a complex order characterized by peptidergic cell groups (vasopressin, gastrin releasing peptide, vasoactive intestinal polypeptide, calbindin, calretinin, corticotrophin releasing factor and enkephalin) that, in most cases, substantially overlap. The retinohypothalamic tract projects bilaterally to virtually all the suprachiasmatic nucleus in both rat (predominantly contralateral) and mouse (symmetric) and its terminal field overlaps that for the geniculohypothalamic tract, but with distinctions visible according to density criteria; neither provides more than sparse innervation of the dorsomedial suprachiasmatic nucleus. In the mouse, the serotonergic terminal field is densest medially and ventrally, but is also distributed elsewhere with varying density. The serotonergic terminal plexus in the rat is densest centromedially and largely, but not completely, overlaps the complete distribution of retinal terminals with density much reduced in the lateral suprachiasmatic nucleus. The locations of vasopressin neurons, retinohypothalamic tract terminals and serotonergic (mouse, rat) or geniculohypothalamic tract (rat) provide evidence for three clear, but not exclusionary, sectors of the suprachiasmatic nucleus. The data, in conjunction with emerging knowledge concerning rhythmically dynamic changes in the size of regions of neuropeptide gene expression in suprachiasmatic nucleus cells, support the view that suprachiasmatic nucleus organization is more complex than a simple "core" and "shell" arrangement. While generalizations about suprachiasmatic nucleus organization can be made with respect to location of cell phenotypes or terminal fields, oversimplification may hinder, rather than facilitate, understanding of suprachiasmatic nucleus structure-function relationships.