Background: Individuals with anxiety disorders display reduced resting-state heart rate variability (HRV), although findings have been contradictory and the role of specific symptoms has been less clear. It is possible that HRV reductions may transcend diagnostic categories, consistent with dimensional-trait models of psychopathology. Here we investigated whether anxiety disorders or symptoms of anxiety, stress, worry and depression are more strongly associated with resting-state HRV. Methods: Resting-state HRV was calculated in participants with clinical anxiety (n = 25) and healthy controls (n = 58). Symptom severity measures of worry, anxiety, stress, and depression were also collected from participants, regardless of diagnosis. Results: Participants who fulfilled DSM-IV criteria for an anxiety disorder displayed diminished HRV, a difference at trend level significance (p = .1, Hedges’ g = -.37, BF10 = .84). High worriers (Total n = 41; n = 22 diagnosed with an anxiety disorder and n = 19 not meeting criteria for any psychopathology) displayed a robust reduction in resting state HRV relative to low worriers (p = .001, Hedges’ g = -.75, BF10 = 28.16). Conclusions: The specific symptom of worry - not the diagnosis of an anxiety disorder - was associated with the most robust reductions in HRV, indicating that HRV may provide a transdiagnostic biomarker of worry. These results enhance understanding of the relationship between the cardiac autonomic nervous system and anxiety psychopathology, providing support for dimensional-trait models consistent with the Research Domain Criteria framework.