The N-methyl-d-aspartate receptor hypofunction is one of the most prevalent models of schizophrenia. For example, healthy subjects treated with uncompetitive N-methyl-d-aspartate receptor antagonists elicit positive, negative, and cognitive-like symptoms of schizophrenia. Patients with anti-N-methyl-d-aspartate receptor encephalitis, which is likely caused by autoantibody-mediated down-regulation of cell surface N-methyl-d-aspartate receptors, often experience psychiatric symptoms similar to schizophrenia initially. However, where and when N-methyl-d-aspartate receptor hypofunction occurs in the brain of schizophrenic patients is poorly understood. Here we review the findings from N-methyl-d-aspartate receptor antagonist and autoantibody models, postmortem studies on N-methyl-d-aspartate receptor subunits, as well as the global and cell-type-specific knockout mouse models of subunit GluN1. We compare various conditional GluN1 knockout mouse strains, focusing on the onset of N-methyl-d-aspartate receptor deletion and on the cortical cell-types. Based on these results, we hypothesize that N-methyl-d-aspartate receptor hypofunction initially occurs in cortical GABAergic neurons during early postnatal development. The resulting GABA neuron maturation deficit may cause reduction of intrinsic excitability and GABA release, leading to disinhibition of pyramidal neurons. The cortical disinhibition in turn could elicit glutamate spillover and subsequent homeostatic down regulation of N-methyl-d-aspartate receptor function in pyramidal neurons in prodromal stage. These two temporally-distinct N-methyl-d-aspartate receptor hypofunctions may be complimentary, as neither alone may not be able to fully explain the entire schizophrenia pathophysiology. Potential underlying mechanisms for N-methyl-d-aspartate receptor hypofunction in cortical GABA neurons are also discussed, based on studies of naturally-occurring N-methyl-d-aspartate receptor antagonists, neuregulin/ErbB4 signaling pathway, and theoretical analysis of excitatory/inhibitory balance.