The use of animal models for the study of psychosis and new treatment development is inadequate in assessing targetpsychotic symptoms because animals lack an ability to uselanguage. Despite this deficiency, new antidopaminergicantipsychotic drugs have still become available. However,even these new antipsychotics, although substantially betterthan the conventional compounds, do not "cure" psychosisor normalize schizophrenic symptoms. The need for newtreatment strategies is apparent. The value of a humanmodel, where language is available to describe targetsymptoms, is clear. Currently, there is an opportunity touse the mild psychotomimetic symptoms induced by aminimal dose of ketamine in normal humans as a model ofpsychosis. The mental symptoms in this model resemble some of the symptoms of schizophrenia, suggesting theadditional possibility that parallel mechanisms of psychosis may occur in schizophrenia and in a ketamine state, creating a potentially viable psychosis model for pathophysiology. This paper includes arguments in support of this human model’s application. Several potential outcome measures that can be used to evaluate potentially novel antipsychotics are described. This model has the potential for identifying novel therapeutics because it does not primarily utilize the dopaminergic system. Further delineation of ketamine pharmacology in humans is pivotal to the eventual application of this ketamine model in drug development. © 1999 American College of Neuropsychopharmacology.