Although many workers have appreciated the striking cytologic and neurochemical similarities of neostriatum, accumbens and olfactory tubercle, a compelling case for regarding these areas as territories in a striatal complex awaited the arguments made by Heimer and his colleagues based on their investigations of connections. A number of recent papers support this viewpoint and extend it with the characterization of three accumbal subterritories: core, shell and rostral pole. The case for separate classifications of systems traversing the accumbens has become more compelling with each study that demonstrates connectional, cytoarchitectural and neurochemical specificity conforming to the boundaries separating the core and its downstream targets from the shell and its projection fields. Furthermore, its apparent composite of core-like and shell-like characteristics distinguishes the rostral pole as yet another unique subterritory. Differences in compartmental organization distinguish the accumbens and neostriatum. The available data are consistent with the periventricular and rostrolateral enkephalin-rich zones being ventralmost parts of the neostriatal patch and matrix compartments, respectively. The accumbal cell cluster compartment, on the other hand, appears to be a separate entity, with connectional and neurochemical features that are dissimilar to both patch and matrix of neostriatum. Boundaries between the accumbens and caudate-putamen remain elusive, and the point of view that such boundaries do not exist but, rather, are represented by "transition zones" must to a large degree reflect the reality. Likewise, it is important to acknowledge that the boundaries between accumbal subterritories are not necessarily distinct or observed faithfully by all of the afferent systems. "Transition zones" appear to be particularly significant organizational features in rostral and lateral parts of the accumbens. Interestingly, histochemically distinct cell clusters tend to be numerous in boundary regions between adjacent territories and subterritories. The predominant organizational pattern appears to be one in which the core, shell and rostral pole engage different forebrain systems that possibly subserve entirely different functions mediated by distantly related mechanisms. In this regard, it is of paramount interest that the processing of information conveyed to the accumbens by diverse cortical and subcortical inputs occurs within distinct and perhaps very different dopaminergic environments in the core, shell and rostral pole (e.g., see Refs 24, 34, 90, 110).