K. Rothermund and D. Wentura (2004) showed how Figure-Ground (FG) asymmetries produce effects on the Implicit Association Task (TAT), independent of associations. Here, the FG account was tested for the robust finding that drinkers show a negative alcohol-IAT effect while being positive on explicit measures. FG asymmetries were manipulated through familiarity of alcohol-IAT target categories and were assessed with visual search tasks. Supporting the FG account, the familiarity manipulation influenced the TAT effect in the expected direction, and the TAT effect correlated with FG asymmetries. Contrary to the FG account, however, the TAT effect was not reversed, and TAT effects were predicted by alcohol use but not by FG asymmetries. Hence, the FG account only partly explains the negative alcohol-IAT effect.