Anxiety and depression cause major detriment to the patient, family, and society - particularly in treatment-resistant (TR) cases, which are highly prevalent. TR prevalence may be due to current diagnoses being based not on biological measures but on symptom lists that suffer from clinical subjectivity, variation in symptom presentation, and comorbidity. Goal-conflict-specific rhythmicity (GCSR) measured using the Stop-Signal Task (SST) may provide the first neural biomarker for an anxiety process and disorder. This GCSR has been validated with selective drugs for anxiety. So, we proposed that GCSR could differ between TR and non-TR individuals and do so differently between those diagnoses normally sensitive to selective anxiolytics and those not. We recorded electroencephalograms (EEG) from 20 TR participants (4 GAD, 5 SAD and 11 MDD) and 24 non-TR participants (4 GAD, 5 SAD and 15 Comorbid GAD/MDD (GMD)) while they performed the SST. There was significant positive GCSR in all groups except the GAD-TR group. GAD-TR lacked GCSR in the low-frequency range. However, TR had little effect in SAD or MDD/GMD populations with apparent increases not decreases. Overall, these results suggest that GAD may occur in two forms: one resulting from excessive GCSR and so being drug sensitive, and the other resulting from some other mechanism and so being TR. In SAD and MDD groups, heightened GCSR could be a consequence rather than the cause, driven by mechanisms that are normally more sensitive to non-selective panicolytic antidepressants.