Face validity in animal models of alcohol abuse and dependence is often at odds with robust demonstrations of ethanol-seeking. This study determined the relative influence of ethanol and a flavorant in maintaining ethanol intake in a nonhuman primate model of "cocktail" drinking. Four-year-old male monkeys were maintained on a 6% ethanol/6% Tang solution made available in daily (M-F) 1-h sessions. Experiments determined the effect of (1) a second daily access session, (2) concurrent presentation of the Tang vehicle, (3) sequential presentation of the vehicle in the first daily session and the ethanol solution in the second session, (4) altering the Tang concentration, (5) altering the ethanol concentration, and (6) removal of the flavorant. Mean daily intake (2.7+/-0.2 g/kg/day) was stable over 7 months. Simultaneous availability of a large, but not a low-moderate, volume of the vehicle reduced ethanol intake by about 50%. Decreasing the concentration of Tang in the first daily session reduced ethanol intake, whereas intake of the standard solution was increased in the second session. Ethanol consumption was decreased by only 27% when the flavorant was removed. In summary, alterations that reduced intake in the first daily session resulted in compensatory increases in ethanol intake in the second session, suggesting that animals sought a specific level of ethanol intake per day. It is concluded that models with excellent face validity (flavored beverages) can produce reliable ethanol intake in patterns that are highly consistent with ethanol-seeking behavior.