This report presents a cognitive measure of drinking expectancies based on semantic priming methodologies, and it investigates differences between high and low drinking restraint undergraduates, Reaction time to an emotion name (e.g., sociable) was measured when that word was preceded by a drinking-related word (e.g., beer) and when preceded by a neutral word (e.g., book). One hundred and four undergraduate students were administered the "differentiated" Restrained Drinking Scale (Ruderman McKirnan, 1984; Collins, George, & Lapp, 1989), a drinking behaviour questionnaire, and completed the priming task. High- and low-restraint drinkers were found to differ significantly in the extent and direction of priming to positive and negative emotional word targets, with the high-restraint drinkers taking longer to recognise positive words, The results suggest that, in the absence of an alcohol pre-load, highly restrained drinkers have weaker positive drinking expectancies than their less restrained peers. The priming procedure overcomes problems of response manipulation that may be present in self-report measures of expectancies.