Monoamine releasers constitute one class of drugs under investigation as candidate medications for the treatment of cocaine abuse. Promising preclinical and clinical results have been obtained with amphetamine, which has high selectivity for releasing dopamine/norepinephrine versus serotonin. However, use of amphetamine as a pharmacotherapy is complicated by its high abuse potential. Recent preclinical studies suggest that nonselective monoamine releasers or serotoninselective releasers have lower abuse liability and may warrant evaluation as alternatives to amphetamine. To address this issue, the present study evaluated the effects of five monoamine releasers in assays of cocaine discrimination and cocaine self-administration in rhesus monkeys. The releasers varied along a continuum from dopamine/norepinephrine-selective to serotonin-selective [m-fluoroamphetamine (PAL-353), methamphetamine, m-methylamphetamine (PAL-314), 1-napthyl-2-aminopropane (PAL-287), fenfluramine]. In drug discrimination studies, rhesus monkeys were trained to discriminate saline from cocaine (0.4 mg/kg i.m.) in a two-key, foodreinforced drug discrimination procedure. Substitution for cocaine was positively associated with selectivity for dopamine/norepinephrine versus serotonin release. In drug selfadministration studies, rhesus monkeys responded for cocaine (0.01 and 0.032 mg/kg/injection) and food (1-g pellets) under a second-order fixed-ratio 2 (variable-ratio 16:S) schedule. In general, monoamine releasers produced dose-dependent and sustained decreases in cocaine self-administration. However, the dopamine/norepinephrine-selective releasers decreased cocaine self-administration with minimal effects on foodmaintained responding, whereas the more serotonin-selective releasers produced nonselective reductions in both cocaineand food-maintained responding. These results are consistent with the conclusion that dopamine/norepinephrine-selective releasers retain cocaine-like abuse-related effects but may also be capable of producing relatively selective reductions in the reinforcing effects of cocaine.Cocaine abuse and dependence constitute a significant public health problem, and reliable pharmacotherapies are not yet available (Mendelson and Mello, 2004;Vocci et al., 2005). One potential approach to the treatment of cocaine dependence has been suggested by the effective treatment of opioid dependence with opioid agonists such as methadone and of tobacco dependence with formulations of nicotine. These medications share pharmacological mechanisms of action with the abused drug and produce some effects in common with the abused drug, and they have been referred to as "agonist" medications (Rothman and Glowa, 1995). Cocaine blocks the reuptake of dopamine, serotonin, and norepinephrine, and the abuse-related effects of cocaine are thought to be mediated primarily by its dopaminergic effects (