How do animals use the same peripheral structures to generate different behavioral responses? Three different neuronal architectures have been proposed to mediate this task: dedicated circuitry; distributed circuitry; and reorganizing circuitry. This review will critically examine the evidence for these different architectures in invertebrate circuits, and then examine the evidence for them in more complex vertebrate circuits. The evidence suggests that these different architectures are unlikely to be found in pure form in most neural circuits, but are useful for guiding the experimental analysis of circuitry.