ABSTRACT Background Implementation intentions have been demonstrated to reduce alcohol use in retrospective self-report measures. It remains unclear, however, whether they can moderate drinking in a challenging naturalistic context. We examined this by studying the effects of implementation intentions on alcohol use in a bar. Methods One hundred and twenty-one participants visiting a local bar were randomized over three conditions: control, motivational trigger, and motivational trigger plus implementation intentions. Those in the motivational trigger group were motivated to reduce drinking, and the combined intervention group made additional alcohol-related implementation intentions. We assessed participants’ pre- and post-intervention breath alcohol concentrations, their post-intervention drinking behavior over one hour, and measures of trait impulsivity and reward sensitivity. To predict post-intervention drinking, multiple hierarchical regression was used. Results Neither the motivational trigger nor the added implementation intentions reduced drinking. Across treatment groups, trait impulsivity predicted alcohol consumption. Conclusions The present study showed that motivational trigger and implementation intentions failed to reduce alcohol use in a naturalistic setting. Potential reasons, implications, and limitations are discussed. This first attempt demonstrates that implementation intentions aimed at moderation are not as easily implemented in a public high-risk context.