Although Behavior Change Support Systems (BCSS) are gaining ground in the field of health in-terventions, we lack an empirically grounded understanding of how the behavior change tech-niques (BCTs) that are implemented in BCSS influence behavioral outcomes. Based on the self-efficacy theory, we conduct narrative interviews to investigate the process along which BCTs ap-plied in wearable activity trackers (WATs) influence users’ perceived self-efficacy and behaviors. We find three patterns that show how WATs’ BCTs feed certain information sources on which users build their self-efficacy beliefs. We identify a positive path (i.e., high self-efficacy, leading to com-pliant behavior) and a negative path (i.e., low self-efficacy, leading to non-compliant behavior) for each of these patterns. Our findings indicate that, under certain circumstances and/or at a cer-tain level of task difficulty, BCTs inflict adverse effects on users’ perceptions of their self-efficacy and their subsequent behavioral responses. Our results provide insights for theory and practice into how BCSS affect perceptions of self-efficacy and behavior change.