Eighteen male social drinkers underwent four training sessions during which they ingested two colour-coded drinks (red or blue, balanced for drink type); one containing alcohol (aliquots of 0.1 g/kg) and the other placebo (aliquots of orangeade). Following the training sessions, subjects were presented with both drinks, and instructed to choose the drink they felt like consuming and to indicate their preference for their chosen drink over the other drink. In addition, they were instructed to consume the first drink but that all subsequent drinks (total of six drinks), offered at 10-min intervals, were optional. A number of trait characteristics were assessed including alcohol outcome expectancies, drinking habits and personality traits. The acute effects of alcohol on mood was also evaluated by comparing subjective ratings following alcohol and placebo during the training sessions. Of the 18 subjects, 12 chose alcohol at least once ('samplers'), whereas six never chose alcohol ('non-samplers'). Over the three sessions, however, alcohol and placebo were chosen equally. When alcohol was chosen, subjects drank significantly more than when placebo was chosen, which may be consistent with a priming effect of drinking alcohol. The amount of alcohol drunk was seen to correlate with the alcohol expectancy factor 'sociability'. Subjective reports of feeling 'alert', 'clear-headed', 'quick-witted', and 'attentive' all showed a main effect of choosing behaviour (i.e. 'samplers'/'non-samplers'). Further analysis indicated that this effect was due to 'samplers' reporting increased subjective ratings of these mood states following the ingestion of alcohol compared to 'non-samplers'. These increased subjective ratings were also positively correlated with the amount of alcohol consumed by the subjects during the choice procedure. No other relationships were found between the amount of alcohol consumed and any of the other state or trait measures. These data suggest that social drinkers who sample alcohol in a laboratory setting can be primed by alcohol to consume more. The results also indicated that the amount drunk was related to the degree to which subjects expected alcohol to increase sociability and to reports of subjective stimulant effects of alcohol (e.g., 'alert', 'clear-headed', 'quick-witted', and 'attentive').