Background: Shame is considered an important factor in the development and maintenance of many psychological disorders, e.g., social anxiety disorder, and an interesting target point for therapeutic intervention. Method: In the present experimental study, we used an online-adopted Autobiographical Emotional Memory Task (AEMT) to induce shame and tested different micro-interventions (self-compassion, cognitive reappraisal, and a control intervention) with respect to their potential to reduce shame intensity. One-hundred-and-fifteen healthy subjects participated in the study and completed a series of self-report questionnaires on self-compassion, shame, and social anxiety. Results: The experimental shame induction was well accepted and successful (with significantly heightened feelings of shame); there were no study drop-outs. There was a significant time*condition interaction, which was due the self-compassion-based intervention resulting in a significantly larger reduction of shame than the control condition (counting fishes). In addition, the main effect of the factor experimental condition was further moderated (enhanced) by trait social anxiety and trait self-compassion. Conclusion: The findings demonstrate the usefulness of online-adopted AEMT for the experimental induction of shame. They suggest that especially self-compassion interventions can be beneficial in alleviating intense shame experiences, which is in accordance with self-compassion theory. Overall, the results are promising in the context of experimental shame research and its potential clinical impacts call for further replication.