Rats implanted with chronic indwelling intravenous catheters and allowed access to a self-administration apparatus learned to self-inject intravenous ethanol. Ethanol concentrations of 0.5, 1.0, and 2.0%, corresponding to a dose/injection of 1, 2, and 4 mg/kg, respectively, were consistently self-injected. Self-injection was not acquired or maintained with ethanol doses of 0.5 or 8 mg/kg/injection. Saline replacement of ethanol reservoirs led to marked increases in lever-pressing response in animals self-injecting 1, 2, and 4 mg/kg ethanol/injection but not with 0.5 or 8 mg/kg/injection. Neurotoxin-induced lesions of dopamine-(DA) containing neurons in nucleus accumbens septi failed to alter the acquisition or maintenance of ethanol self-injection. Pretreatment with haloperidol (0.05 and 0.1 mg/kg, SC) failed to alter hourly or daily self-injection rates. On the other hand, p-chlorophenylalanine pretreatment increased, while fluoxetine (2.5 and 5.0 mg/kg) administration significantly reduced, self-injected intravenous ethanol. These data suggest that ethanol is self-injected by the rat in a narrow dose range and that 5-hydroxytryptamine (5-HT), but not DA-containing neurons, subserves some function in the reinforcing or aversive affects of ethanol.