This paper discusses the role of inhibition in neural and cognitive control, in particular its role in selective processes in perception and action. We first review neuroanatomical and physiological evidence that cortical control is mediated by a variety of local-circuit inhibitory neurons, distributed throughout all layers and areas of the cortex. We then consider the use of inhibition in control processes in behavioral neural network models, focusing on two areas: selective attention and sequential action. Relations between the architecture and dynamics of these models and relevant neurological findings are discussed. We conclude that a full understanding of inhibitory control of mental functioning will require the integration of data from both the behavioral and neural levels, and that formal neural network models can play an important role in bridging this epistemological divide.