Recently, our laboratory has demonstrated that predatory attack behavior in the cat, elicited by electrical stimulation of the lateral hypothalamus, is suppressed following activation of the region of the medial hypothalamus from which defensive rage behavior is elicited [Han, Y., Shaikh, M.B., Siegel, A., Medical amygdaloid suppression of predatory attack behavior in the cat: II. Role of a GABAergic pathway from the medial to the lateral hypothalamus, Brain Res., 716 (1996) 72-83.]. The mechanism for this suppression is a direct GABAergic projection from the medial to lateral hypothalamus. The present study tested the hypothesis that the inhibitory relationship between these two regions of hypothalamus is reciprocal, namely, that a GABAergic neuron, which also projects from the lateral to medial hypothalamus, serves to suppress defensive rage elicited from the medial hypothalamus. Monopolar stimulating electrodes were implanted into lateral hypothalamic sites from which predatory attack behavior was elicited. In addition, cannula-electrodes were implanted into the medial hypothalamus for elicitation of defensive rage behavior and for microinjections of GABA compounds. Initially, in the absence of drug administration, the effects of dual stimulation of the lateral and medial hypothalamus upon response latencies were compared with those following single stimulation of the medial hypothalamus alone. Dual stimulation significantly (p<0.01) suppressed defensive rage behavior elicited from the medial hypothalamus. Then, administration of the GABAA receptor antagonist, bicuculline (10-60 pmol), into medial hypothalamic sites from which defensive rage was elicited blocked the suppressive effects of lateral hypothalamic stimulation. The GABAA receptor agonist, muscimol (0.3-30 pmol), microinjected into the medial hypothalamus, suppressed defensive rage elicited by single stimulation of the medial hypothalamus in a dose dependent manner. These suppressive effects of muscimol upon defensive rage were blocked following pretreatment with bicuculline (60 pmol). Administration of muscimol into adjoining regions of the lateral hypothalamus had no effect upon defensive rage, indicating its site specificity. Bicuculline (60 pmol) delivery into the medial hypothalamus had no effect upon defensive rage, suggesting the, presence of a phasic rather than tonic mechanism. A combination of immunocytochemical and retro grade tracing procedures were then employed to determine the origin of the putative GABAergic pathway projecting to the medial hypothalamus. In this experiment, the retrograde tracer, Fluoro-Gold (8%, 0.5 microl), was microinjected through a cannula-electrode in the medial hypothalamus from which defensive rage had been elicited. Following survival periods of 5-6 days, cats were perfused with 4% paraformaldehyde and brain tissue was processed for immunocytochemical staining of GABA neurons. Retrogradely labeled, immunopositively labeled, as well as Fluoro-Gold and GABA labeled cells, were identified in the lateral hypothalamus. Each type of neuron was distributed over wide regions of the lateral hypothalamus, extending from the area immediately caudal to the optic chiasm to the level of the posterior hypothalamus. Together, the behavioral pharmacological and anatomical data provide evidence of a direct inhibitory projection from the lateral to medial hypothalamus whose functions are mediated by GABAA receptors. When coupled with our previous findings, these results reveal the presence of reciprocal GABAergic inhibitory pathways between the medial and lateral hypothalamus. The findings suggest that functions associated with either the lateral or medial hypothalamus, but not both, can be activated at a given time.