Pharmacological results obtained in animals tested in approach/avoidance conflict situations have led to the suggestion that 5-HT enhances anxiety by acting on forebrain structures. In contrast, results with intracerebral drug injection associated with aversive electrical brain stimulation indicate that 5-HT inhibits aversion in the dorsal periaqueductal gray (DPAG). To reconcile this evidence, it has been suggested that 5-HT may enhance conditioned fear in the amygdala while inhibiting innate fear in the DPAG. To test this hypothesis, we used three drug treatments known to increase the release of 5-HT from terminals of the dorsal raphe nucleus (DR): (1) intra-DR microinjection of the benzodiazepine inverse agonist FG 7142, (2) intra-DR microinjection of the excitatory amino acid kainic acid and (3) intraperitoneal injection of the 5-HT releaser and uptake blocker D-fenfluramine. All drug treatments enhanced inhibitory avoidance (learned fear) in the elevated T-maze, a new animal model of anxiety. Intra-raphe kainate and D-fenfluramine also decreased one-way escape (innate fear) in the T-maze. In contrast, reduction of 5-HT release by intra-DR injection of 8-OH-DPAT impaired inhibitory avoidance without affecting one-way escape. Overall, these results agree with the above hypothesis. Clinical implications are discussed, especially with regard to panic disorder.