Alcohol 'cognitions' were explored using an implicit methodology [Stacy, Leigh and Weingardt, 1994]. In Study 1, an Associations Questionnaire was developed with young adult undergraduates (median=20 years) comprising culturally available (i.e., high-frequency occurrence) and idiosyncratic (i.e., low-frequency occurrence) positive and negative alcohol consumption outcomes and positive and negative outcomes of behaviors-other-than-alcohol consumption. In Study 2, the relationship was explored between the alcohol consumption of young adult undergraduates (median AGE=19 years) and implicit alcohol-related associations made through the Associations Questionnaire. A significant positive relationship was found between consumption and positive and negative culturally available (experimental) outcomes but not for the other two types of (control) outcome. In Study 3, the relationship between alcohol consumption and alcohol-related associations was explored in mature adults (median=45 years) with many more years' drinking experience. The results of Study 1 were replicated except that a significant positive relationship between consumption and some control outcomes was now found—e.g., negative outcomes of what where hitherto behaviors-other-than-alcohol consumption. In Study 4, however, using the same techniques that were used to develop the Associations Questionnaire in Study 1, an extended set of negative alcohol consumption outcomes was found in mature adults (median=44 years) that included a proportion of the negative outcomes of behaviors-other-than-alcohol consumption that had served as controls in Studies 2 and 3. A reanalysis of the data from Study 3, with such items removed from the controls and designated 'new' negative alcohol consumption outcomes, showed a positive relationship between consumption and implicit alcohol-related associations made through these 'new' items of the Associations Questionnaire. The changing profile of associations with negative outcomes of consumption is discussed and related to negative expectancy research and drinking restraint. Yes Yes