Alcohol expectancies are theorized to operate through associative memory networks. These networks are thought to differ on the basis of individual experience (direct and vicarious) with alcohol. Free-associate probabilities have been used in cognitive psychology as a metric of the associative strength of words to other words; this method has been used to establish the relationships within a semantic memory network. Participants from a large college sample were asked to free associate to the phrase" Alcohol makes me______" They were also asked about their quantity of alcohol consumption. Results showed that specific responses were given with different probabilities by individuals who drank at different levels. The heaviest drinkers tended to have more positive and arousing responses than did lighter drinkers, who tended to have more negative and sedating responses. These results underscore the need to take into account relevant individual differences in behavior and experience when characterizing semantic networks.(PsycInfo Database Record (c) 2023 APA, all rights reserved)