Objective: Threat-related attention bias figures prominently in contemporary accounts of the maintenance of anxiety disorders, yet longitudinal intervention research relating attention bias to anxiety symptom severity is limited. Capitalizing on recent advances in the conceptualization and measurement of attention bias, we aimed to examine the relation between attention bias, indexed using trial-level bias scores (TLBSs) to quantify temporal dynamics reflecting dysregulation of attentional processing of threat (as opposed to aggregated mean bias scores) and social anxiety symptom severity over the course of cognitive-behavioral therapy (CBT) and 1-month follow-up. Method: Adults with social anxiety disorder (N = 39) assigned to either yohimbine- or placebo-augmented CBT completed measures of attention bias and social anxiety symptom severity weekly throughout CBT (5 sessions) and at 1-week and 1-month posttreatment. Results: TLBSs of attention bias temporal dynamics showed stronger psychometric properties than mean aggregated scores and were highly interrelated, in line with within-subject temporal variability fluctuating in time between attentional overengagement and strategic avoidance from threat. Attention bias toward threat and temporal variability in attention bias (i.e., attentional dysregulation), but not attention bias away from threat, significantly reduced over the course of CBT. Cross-lag analyses revealed no evidence of a causal relation between reductions in attentional dysregulation leading to symptom severity reduction, or vice versa. Observed relations did not vary as a function of time. Conclusions: We found no evidence for attentional dysregulation as a causal mechanism for symptom reduction in CBT for social anxiety disorders. Implications for future research are discussed. What is the public health significance of this article? This study suggests that attentional dysregulation may not be a mechanism for change in cognitive-behavioral therapy, an effective treatment for social anxiety disorder. Though data are still preliminary, our finding that cognitive-behavioral therapy did not lead to changes in attentional avoidance leaves open the possibility that targeting attentional avoidance alongside cognitive-behavioral therapy may enhance its efficacy.