Abstract Rationale Recently, experimental paradigms have been developed to strengthen automatic avoidance or inhibitory responses for smoking cues. However, these procedures have not yet been directly compared regarding their effectiveness and mechanisms of action. Objective This study compared the effects of avoidance vs. inhibitory training as an add-on to a brief smoking cessation intervention. The standard Approach-Avoidance-Task (AAT) was adapted for both training types and control conditions. Methods One hundred twenty-four smokers attended behavioral counseling for smoking cessation and were thereafter randomized to one of four training conditions: avoidance-AAT, sham-avoidance-AAT, inhibition-AAT, sham-inhibition-AAT. During a 2-week training period including five training sessions, smokers in the avoidance-AAT trained to implicitly avoid all smoking-related cues, while smokers in the inhibition-AAT trained to implicitly inhibit behavioral response to smoking cues. During sham training, no such contingencies appeared. Self-report and behavioral data were assessed before and after training. Cigarette smoking and nicotine dependence were also assessed at 4- and 12-week follow-ups. Results At posttest, avoidance training was more effective in reducing daily smoking than inhibition training. However, this difference was no longer evident in follow-up assessments. All training conditions improved other smoking- and health-related outcomes. Neither training changed smoking-related approach biases or associations, but approach biases for smoking-unrelated pictures increased and Stroop interference decreased in all conditions. Smoking devaluation was also comparable in all groups. Conclusions Avoidance training might be slightly more effective in reducing smoking than inhibitory training. Overall, however, all four training types yielded equivalent therapy and training effects. Hence, a clear preference for one type of training remains premature.