Clinical and preclinical data suggest a link between serotonin [5-hydroxytryptamine (5-HT)] function and certain psychopathologic dimensions of anxiety disorders. Antidepressants consistently have been found to exert a favorable effect in anxiety disorders, particularly panic disorders. Clinical studies with 5-HT-selective drugs have shown that 5-HT neurons may comprise the site at which anxiolytic drugs exert a significant proportion of their action. Thus, fluvoxamine, a selective 5-HT uptake inhibitor, but not maprotiline, a selective noradrenaline uptake inhibitor, was found to be efficacious in panic disorder. The clinical effect of fluvoxamine revealed a noteworthy time course. After an initial increase in anxiety, improvement was attained gradually. On the basis of this finding, we tentatively hypothesized that stimulation of the 5-HT receptors, resulting from uptake inhibition, would worsen the condition of the patient, while down-regulation of the 5-HT receptors, resulting from chronic treatment, would account for the clinical efficacy. Thus, we performed a study in which ritanserin, a putative 5-HT2 antagonist, was compared with fluvoxamine. Ritanserin was found to be ineffective in the treatment of panic disorder symptoms, suggesting that 5-HT2 receptors may not be critically involved in the mechanism underlying the anxiolytic activity of 5-HT uptake inhibitors. It would seem, therefore, that other 5-HT-receptor subtypes, e.g., 5-HT1, may be implicated in this effect. Recent studies with selective 5-HT1 agonists support this hypothesis.