A neuropsychological theory of the positive symptoms of schizophrenia proposed by J.A. Gray et al. is reviewed in light of subsequent evidence from studies of latent inhibition in clinical populations, healthy volunteers, and rats. It is clear that disrupted latent inhibition is associated with psychosis, but it is uncertain whether this is a state or a trait marker. Much evidence indicates that a similar disruption in latent inhibition can be provoked in rats by excess release of dopamine in the nucleus accumbens, and conversely, that potentiation of latent inhibition occurs when dopaminergic transmission is blocked in this structure. The projection from the hippocampal system to the nucleus accumbens also plays a role in latent inhibition. The theory, therefore, is broadly supported by recent findings. The resulting model of schizophrenia is discussed in relation to the contents of consciousness, positive psychotic symptoms, and alternative theories.