Monoamine releasers with varying selectivity for dopamine (DA)/norepinephrine and serotonin (5-HT) release are potential treatment medications for cocaine abuse. Although DA-selective monoamine releasers effectively reduce cocaine abuse, their clinical usefulness is limited by abuse liability. It is hypothesized that increasing 5-HT neurotransmission may reduce the abuse-related effects of DA releasers, but the optimal DA:5-HT release ratio remains to be determined. This study in rhesus monkeys compared the effects of two compounds with differing potency for 5-HT release. Methcathinone and 3-Cl-methcathinone (PAL-434) have equal potency for DA release, but PAL-434 has 10-fold higher potency for 5-HT release. In drug discrimination studies, monkeys were trained to discriminate cocaine (0.4 mg/kg i.m.) from saline in a two-key, food-reinforced procedure. In drug self-administration studies, a separate group of monkeys was trained to respond for cocaine [0.01 mg/kg/injection (inj)] and food (1 g pellets) under a second order schedule of reinforcement [FR2(VR16:S)]. When responding was stable, methcathinone (0.1–0.56 mg/kg.h i.v.) or PAL-434 (0.32–1.8 mg/kg.h i.v.) was administered chronically (one injection every 20 min for 23 h/d) for 7–10 d. In discrimination studies, both compounds dose-dependently increased cocaine-like responding but with different potencies (cocaine=methcathinone >PAL-434). Chronic treatment with methcathinone or PAL-434 dose-dependently and selectively reduced cocaine self-administration. PAL-434 was about 4-fold and methcathinone about 1.6-fold more potent at decreasing cocaine- over food-maintained responding. These data suggest that compounds with moderate selectivity for DA vs. 5-HT release (8–15-fold) may be effective for the treatment of cocaine dependence.