Combined dopamine (DA) and 5-hydroxytryptamine (5-HT) releasers such as phentermine (PHEN) and fenfluramine (FEN) are reported, in open label studies, to reduce craving for alcohol and cocaine and to prevent relapse. The objective of the studies reported here was to assess the actions of these agents alone and in combination in various animal models of drug addiction. Study 1. in vivo microdialysis experiments demonstrate that these agents preferentially release mesolimbic DA (PHEN) and 5-HT (FEN). Patients who relapse and use cocaine while taking these medications report diminished cocaine-like subjective effects. Microdialysis experiments were performed in awake rats, and dialysate samples were analyzed for DA and 5-HT. PHEN (1 mg/kg, intravenously (i.v.)) elevated DA (2-3-fold) for over 1.5 hr. Administration of cocaine (3 mg/kg, i.v.) increased DA 6-fold in saline-treated rats, but only 3-fold in PHEN-treated rats. PHEN did not reduce cocaine-induced increases in 5-HT. Study 2. These agents were assessed in a mouse model of cocaine-conditioned motoric activity (CCMA). Pretreatment with non-activating doses of PHEN (4.6 mg/kg, intraperitoneally (i.p.)) enhanced CCMA, whereas non-depressing doses of FEN (0.1 mg/kg, i.p.) did not alter CCMA or the PHEN-induced increase in CCMA. In contrast, sub-effective doses of FEN reduced CCMA stereotypy-like locomotion, whereas sub-effective doses of PHEN were without effect. PHEN reversed the FEN-induced increase in CCMA stereotypy-like locomotion. Study 3. PHEN and FEN were assessed in the conditioned place preference model. FEN produced marked aversions for an environment previously associated with its administration and the minimum dose producing this effect was 3.0 mg/kg. In contrast, administration of PHEN, amphetamine (1.0-3.0 mg/kg) or morphine (3.0-5.0 mg/kg) produced dose-related preferences for the drug-paired place. However, the magnitude of the response to PHEN was less than that produced by the other prototypic drugs of abuse. In rats that received FEN (0.3 or 3.0 mg/kg) in combination with PHEN (3.0 mg/kg), the conditioned rewarding effects of PHEN were abolished. These data demonstrate that the rewarding effects of PHEN can be conditioned to stimuli previously associated with its administration. However, the conditioned response to this agent is less then that produced by prototypic drugs of abuse. The finding that PHEN-induced place preferences were attenuated by doses of FEN demonstrates that the combination of FEN/PHEN is devoid of motivational effects. The preclinical data obtained with PHEN/FEN in various models of drug provide a strong rationale for pursuing controlled clinical trials in humans with agents that act via a similar mechanism of action.