Findings presented in this paper from pharmacogenetic studies of oral ethanol self-administration suggest a correlation between ethanol preference and self-administration and indicate that there are important genetic as well as environmental determinants of ethanol reinforced behavior. AA (alcohol accepting) and ANA (alcohol nonaccepting) rats, animals bred selectively for differential ethanol preference, showed large differences in operant responding for ethanol. AA rats drank significantly more ethanol than water, and their intake varied as a function of ethanol concentration. Intake of water and ethanol solutions did not differ in the ANA rats. In two inbred strains of rats, F344 and LEWIS, ethanol maintained higher response rates and was consumed in larger volumes than the water vehicle. In a third series of studies, C57BL/6J mice, which exhibit high ethanol preference and low sensitivity, readily self-administered ethanol in an operant situation. Conversely, BALB/cJ mice, which exhibit low preference and high sensitivity, were not positively reinforced by ethanol. The results demonstrate the experimental control possible by the utilization of genetically defined animals, even when complex learned behavioral sequences are being measured, and indicate that genotype and environment interact in a complex but definable way to determine the degree to which ethanol comes to function as a positive reinforcer.