Excessive alcohol drinking, particularly among college students, is a major health concern worldwide. The implicit associations between alcohol-related concepts and affective attributes have been repeatedly postulated as a reliable predictor of these drinking behaviors. The Implicit Association Test (IAT) is considered one of the most reliable tasks for measuring these associations and their impact on actual alcohol consumption. However, the majority of these tests used verbal materials as stimuli, thus being unadapted to some categories of participants. The present study aims to develop a new IAT, using pictures exclusively as stimuli, to provide a cross-cultural and language-independent evaluation of implicit associations that is more closely related to real-life drinking contexts. Sixty-five undergraduate young adults took part in this study. A new visual IAT was used to measure the implicit association between alcohol cues and alcohol-related positive attributes. Pictorial stimuli, previously validated, were used to represent both target (alcohol vs. soft drinks) and attribute (positive vs. neutral affective states) categories in seven successive experimental blocks. The IAT was followed by self-reported measures of explicit alcohol-related expectancies and alcohol consumption. The new IAT highlighted significant implicit associations between positively valenced and alcohol-related representations conveyed by pictures, with good internal consistency, thus proving its validity and reliability. Importantly, regression analyses showed that these implicit associations are a strong predictor of self-reported alcohol consumption. This visual IAT further underscores that positive implicit associations with alcohol constitute an important factor in predicting effective alcohol-related behaviors and offers a more ecological and cross-cultural way to test these associations in non-alcohol-dependent populations. Moreover, this version of the IAT might be implemented in prevention and prophylactic programs.