There is growing evidence for the involvement of glutamatergic abnormalities in schizophrenia. Uncompetitive NMDA receptor (NMDAR) antagonists induce effects closely resembling both the positive and negative symptoms of schizophrenia; candidate risk genes for schizophrenia converge on the NMDAR expressing synapse; and a recent trial of a drug with direct action at metabotropic glutamate autoreceptors has demonstrated equivalent efficacy to olanzapine in patients with chronic schizophrenia. Imaging the glutamate system in humans in vivo poses a number of difficulties, and has progressed slowly in comparison to the relative ease of dopamine imaging. Indirect imaging of the glutamate system is possible using pharmacological challenges targeting the glutamate system combined with fMRI, PET or SPECT imaging. There are two methods of directly estimating glutamatergic neurotransmission in living patients using neuroimaging at present: [123I]CNS-1261 SPECT (measuring NMDAR binding), and proton magnetic resonance spectroscopy (MRS) of glutamate and glutamine. Both methods have yielded some intriguing insights into glutamatergic abnormalities and their relevance to psychotic symptoms. In this review, the glutamate hypothesis of schizophrenia, and its relationship to current findings in glutamate imaging in psychosis to this hypothesis will be discussed. The possibility of developing new drugs for schizophrenia in light of these findings will then be considered.