Anxiety is one of the most prevalent mental health problems around the world. Despite a number of widely available interventions, it can take weeks or months to see effects, and nearly half of individuals may not respond. In an effort to better understand response rates, a large body of evidence indicates the most consistent predictor of treatment outcomes is activity in the anterior cingulate cortex (ACC). Although activity in ACC can be measured by medial frontal theta event related potentials (ERPs) at a finer temporal resolution, these neurophysiological components have not been evaluated as predictors of treatment response. There is also a lack of research on the functional networks associated with ACC treatment prediction, despite implications for prefrontal engagement of cognitive control processes. The present study aimed to examine these gaps in the literature by using task-based electroencephalography (EEG) and medial frontal theta negativities (MFTN) as predictors of anxiety sensitivity treatment response. Using amplitude as well as functional connectivity measures (i.e., inter-channel phase synchrony), baseline MFTN (i.e., Theta-FN, Theta-N2) were assessed as predictors of treatment response at mid-treatment, 1-week post treatment, and 6 months post treatment. Subjects underwent a baseline EEG before completing three sessions of a computerized cognitive behavioral intervention. Contrary to the hypothesis, findings revealed MFTN amplitude did not predict treatment response. However, medial to lateral prefrontal theta phase synchrony demonstrated significant prediction effects, such that lower phase synchrony was associated with greater symptom improvement at mid-treatment, 1-week post treatment, and 6 months post treatment. This effect was specific to certain task conditions (i.e., gain feedback and go stimuli), as well as to the combined anxiety and depression treatment group. Results demonstrated accuracy and consistency of treatment prediction, as well as incremental validity after controlling for self-report measures. Finally, results provide additional support for a convergent medial frontal theta process, and suggest that low engagement of regulatory and proactive control mechanisms may be predictive of better response to cognitive behavioral interventions. This work represents a novel finding that may contribute to the improvement in treatment efficacy by serving as a target for future interventions and individualized treatment selection.