We have shown a decrease in cortical serotonin(2A) receptors using tissue sections, but not with washed membranes, from the same cohort of subjects. These discrepant findings led us to determine if we could obtain similar results using samples from the same tissue block. Our studies used single-point saturation analyses to estimate the total number of [(3)H]ketanserin binding sites in tissue sections, crude homogenate, membrane-enriched and cytosol-enriched tissue samples from Brodmann's area 9. There were significant decreases in the levels of [(3)H]ketanserin binding using tissue sections (mean+/-SD: 38+/-16 vs. 56+/-16 fmol/mg ETE; p=0.008) and crude tissue homogenates (131+/-53 vs. 168+/-38 fmol/mg protein; p<0.05) from subjects with schizophrenia compared to that in controls. By contrast, there was no significant difference in radioligand binding to the membrane-enriched (155+/-95 vs. 145+/-48 fmol/mg protein; p=0.72) or cytosol-enriched (8.6+/-14 vs. 7.5+/-10 mol/mg protein; p=0.85) tissue fraction. Significantly, adding 10(-5) M risperidone or chlorpromazine, as surrogates for residual antipsychotic drugs in the CNS, to crude homogenate from control subjects did not alter [(3)H]ketanserin binding. Our data therefore is consistent with the hypothesis that apparent decreases in serotonin(2A) receptors in schizophrenia are due to altered levels of a regulatory factor(s) that modulates the binding of ligands to the serotonin(2A) receptor and that separating the membrane and cytosol removes this regulatory control.