Attentional bias variability is related to alcohol abuse. Of potential use for studying variability is the anticipatory attentional bias: Bias due to the locations of predictively-cued rather than already-presented stimuli. The hypothesis was tested that conflicting automatic associations are related to attentional bias variability. Further, relationships were explored between anticipatory biases and individual differences related to alcohol use. 74 social drinkers performed a cued Visual Probe Task and univalent Single-Target Implicit Associations Tasks. Questionnaires were completed on risky drinking, craving, and motivations to drink or refrain from drinking. Ambiguity was related to attentional bias variability at the 800 ms Cue-Stimulus Interval. Further, a bias related to craving and risky drinking was found at the 400 ms Cue-Stimulus Interval. Thus, the selection of attentional responses was biased by predicted locations of expected salient stimuli. The results support a role of conflicting associations in attentional bias variability.