Maladaptability, particularly of autonomic activity, is described as a central component of vulnerability-stress-models for the pathogenesis of psychotic symptoms. Investigating heart rate variability (HRV) as an index of autonomic adaptability is thus likely to improve our understanding of psychosis. In clinically vulnerable groups for psychosis, it is unclear whether maladaptability is already evident. Moreover, to investigate specificity, direct comparisons to other mental disorders are required. In the present study, we analyzed 3鈥痬in of resting-state heart rate, HRV, and negative affect in 130 participants; consisting of participants with psychotic disorders (PSY; n鈥?鈥?4), clinical high-risk for psychosis (CHR; n鈥?鈥?2), anxiety disorders (anxiety controls, AC; n鈥?鈥?9) and healthy controls (HC; n鈥?鈥?5). ANCOVAs controlling for age revealed significant group differences for both investigated vagal HRV parameters, which were reduced in PSY compared to HC. The high-frequency domain HRV in PSY was also lower than in CHR and - in a non-significant trend - than in AC. Also, ANOVAs for heart rate and negative affect revealed significant increases in PSY compared to HC. Exploratory analyses of medication effects showed moderate dosage associations with heart rate and high-frequency HRV. Thus, in the present study, the activity of the autonomic nervous system was altered in psychosis but not in an at-risk group. A potential specificity of the effect can be speculated in contrast to anxiety disorders. Future studies should investigate the predictive value of HRV for increased stress-sensitivity or transition to clinical symptoms as well as the implications for daily threat perception and symptom maintenance.