Conditioned taste aversions induced by pairing the consumption of saccharin with an amphetamine injection are attenuated in rats with depletion of central catecholamines caused by intraventricular administration of 6-hydroxydopamine (6-OHDA). The hypothesis that dopamine (DA) depletion is responsible for this effect was tested. The reduction in conditioning caused by intraventricular 6-OHDA could not be duplicated either with injections of 6-OHDA into the substantia nigra (Experiment 1) or with intraventricular 6-OHDA injections in animals pretreated with desmethylimipramine (Experiment 2). Both treatments, however, produced large depletions of telencephalic DA. 6-Hydroxydopa infusions caused a preferential loss of telencephalic norepinephrine (NE) but also failed to alter taste aversion learning. It is concluded that the effect of intraventricular 6-OHDA on amphetamine-induced aversions was the result of depletion of both NE and DA. In a third experiment the generality of the effect was examined by pairing saccharin consumption with injections of the amphetamine congener fenfluramine. Depletion of both NE and DA failed to alter fenfluramine-induced aversions. Infusion of 6-OHDA into the substantia nigra, however, retarded the extinction of such an aversion. Evidence is discussed for a peripheral site of action for fenfluramine in the conditioned aversion paradigm.Many early laboratory studies of condi-mans, and all were found to be effective in tioned taste aversion (CTA) learning paired the CTA paradigm. Thus, it appeared that a taste stimulus with a treatment that in-any treatment capable of producing nausea duced gastric distress. This single pairing or gastrointestinal distress could be used to was sufficient to cause a suppression of the produce an aversion (Garcia & Koelling, consumption of that taste stimulus on later 1966). Recent data challenge this position, presentations. X-rays and a variety of Several highly toxic rodenticides do not indrugs, including cyclophosphamide, apo-duce CTA learning, although they produce morphine, and lithium chloride were among obvious signs of illness (Nachman & Hartley, the toxicosis-inducing agents tested (Garcia 1975). Furthermore, treatments that pro-