Deficient inhibitory neurotransmission of gamma-aminobutyric acid (GABA) has been implicated in the pathophysiology of schizophrenia based on postmortem studies. However, in vivo studies have shown predominantly negative or conflicting results. The goal of this study was to better characterize possible changes of the regional GABA(A)-benzodiazepine receptor distribution volume (BZR V3-p) in schizophrenia in vivo, using a larger sample size than previous studies. Single photon emission computed tomography (SPECT) with [123I]iomazenil was used with a constant infusion paradigm to measure the BZR V3-p under sustained radiotracer equilibrium conditions. Twenty-five patients with schizophrenia and 24 matched healthy control subjects were studied. Positive and Negative Syndrome Scale (PANSS) ratings were done in all subjects. Statistical parametric mapping (SPM) 96 was used to compare patients and control subjects as well as to study the relationship between SPECT results and composite PANSS scores based on two factorial models: the pentagonal model (positive, negative, dysphoric mood, activation, and autistic preoccupation factors) and the taxometric model (disorganized dimension). On the basis of 'absolute' values of V3-p with no normalization for total brain uptake, the schizophrenic patients showed no significant differences in BZR levels compared to the healthy control subjects. With a global normalization procedure, which is more sensitive to relative regional differences in activity, BZR V3-p was significantly decreased in the patients in the left precentral gyrus (BA 6). The relative BZR V3-p showed a significant positive correlation with duration of illness in the superior occipital gyri (BA 19). No significant correlations were observed between either absolute or relative BZR V3-p and either age or any of the composite PANSS scores based on any of the two factorial models in either patients or control subjects. No significant differences were observed between cigarette smoking vs. non-smoking patients, nor between the patients on atypical antipsychotics vs. on typical antipsychotics vs. not on any antipsychotics. In general, no significant differences in BZR V3-p were observed between patients and control subjects, except for a decrease in relative BZR V3-p in the left precentral gyrus. Grey matter atrophy is unlikely to be the cause for this decrease. However, we could not exclude that possibility. The positive correlation with duration of illness might reflect the relative preservation of neurons expressing BZR in the superior occipital gyri as compared to other cortical brain regions in schizophrenia.