Inasmuch as the presence of endogenous gamma-aminobutyric acid (GABA) may affect benzodiazepine binding to tissue sections in autoradiographic studies, a protocol designed to check for this influence has been investigated. [3H]Flunitrazepam (1 nM) was used to label benzodiazepine receptors for autoradiographic localization. Bicuculline was added to the incubation medium of an additional set of tissue sections to antagonize any potential effect of endogenous GABA. Binding in these sections was compared to that occurring in another set in which excess GABA was added to "create" further GABA enhancement. Binding also was compared to adjacent sections which were treated similarly but also preincubated in distilled-deionized water to burst the cells by osmotic shock and eliminate endogenous GABA, thereby preventing any effect on benzodiazepine binding. The results indicated that endogenous GABA is indeed present in the slide-mounted tissue sections and is affecting benzodiazepine receptor binding differentially in various regions of the brain depending on the density of GABAergic innervation. Scatchard analysis of saturation data demonstrated that the alteration in benzodiazepine binding due to GABA was a result of a change in the affinity rather than number of receptors present. These experiments have been compared to the binding of the imidazodiazepine, [3H] Ro15-1788. We also show that the treatments affect endogenous GABA and not the receptors themselves. This suggests strongly that, when using a single nonsaturating concentration of radiolabeled benzodiazepine antagonist, autoradiographic studies to date may have been subject to erroneous interpretation due to the differential effects of endogenous GABA on benzodiazepine binding (increased affinity).(ABSTRACT TRUNCATED AT 250 WORDS)