Background: Recent theoretical models emphasize the role of impulsive processes in alcohol addiction, which can be retrained with computerized Cognitive Bias Modification (CBM) training. In this study, the focus is on action tendencies that are activated relatively automatically. Objective: The aim of the study is to examine the effectiveness of online CBM Alcohol Avoidance Training using an adapted Approach-Avoidance Task as a supplement to treatment as usual (TAU) in an outpatient treatment setting. Methods: The effectiveness of 8 online sessions of CBM Alcohol Avoidance Training added to TAU is tested in a double-blind, randomized controlled trial with pre- and postassessments, plus follow-up assessments after 3 and 6 months. Participants are adult patients (age 18 years or over) currently following Web-based or face-to-face TAU to reduce or stop drinking. These patients are randomly assigned to a CBM Alcohol Avoidance or a placebo training. The primary outcome measure is a reduction in alcohol consumption. We hypothesize that TAU + CBM will result in up to a 13-percentage point incremental effect in the number of patients reaching the safe drinking guidelines compared to TAU + placebo CBM. Secondary outcome measures include an improvement in health status and a decrease in depression, anxiety, stress, and possible mediation by the change in approach bias. Finally, patients’ adherence, acceptability, and credibility will be examined. Results: The trial was funded in 2014 and is currently in the active participant recruitment phase (since May 2015). Enrolment will be completed in 2019. First results are expected to be submitted for publication in 2020. Conclusions: The main purpose of this study is to increase our knowledge about the added value of online Alcohol Avoidance Training as a supplement to TAU in an outpatient treatment setting. If the added effectiveness of the training is proven, the next step could be to incorporate the intervention into current treatment.