Narcolepsy is currently treated with anti-depressants to control REM-related symptoms such as cataplexy and with amphetamine-like stimulants for the management of sleepiness. Both stimulant and antidepressant drugs presynaptically enhance monoaminergic transmission but both classes of compounds lack pharmacological specificity. In order to determine which monoamine is selectively involved in the therapeutic effect of these compounds, we examined the effects of selective monoamine uptake inhibitors and release enhancers on cataplexy using a canine model of the human disorder. A total of 14 compounds acting on the adrenergic (desipramine, nisoxetine, nortriptyline, tomoxetine, viloxazine), serotoninergic (fenfluramine, fluoxetine, indalpine, paroxetine, zimelidine) and dopaminergic (amfonelic acid, amineptine, bupropion, GBR 12909) systems were tested. Some additional compounds interesting clinically but with less pharmacological selectivity, i.e., cocaine, dextroamphetamine, methylphenidate, nomifensine and pemoline, were also included in the study. All compounds affecting noradrenergic transmission completely suppressed canine cataplexy at low doses in all dogs tested, whereas compounds which predominantly modified serotoninergic and dopaminergic transmission were either inactive or partially active at high doses. Our results demonstrate the preferential involvement of adrenergic systems in the control of cataplexy and, presumably, REM sleep atonia. Our findings also demonstrate that canine narcolepsy is a useful tool in assessing the pharmacological specificity of antidepressant drugs.