Heart rate variability (HRV) monitoring is increasingly applied in the realm of mental disorders; however, it remains a subject of controversy. This umbrella review summarizes HRV differences between individuals with mental disorders and healthy controls (HCs), as well as changes in HRV before and after treatment in patients with mental disorders. A literature search was conducted using Medline, PubMed, Embase, and the Cochrane Database. Meta-analyses on HRV changes in patients with mental disorders, as well as meta-analyses comparing HCs and patients with mental disorders were included. We computed the summary effect size using random effects models, along with 95% confidence and prediction intervals. We assessed heterogeneity, p value of the largest study, excess significance bias, and small-study effects. Evidence levels were classified as convincing, highly suggestive, suggestive, weak, or not significant. Twenty-one systematic reviews on HRV, covering 19 mental disorders (53 meta-analyses) and 8 treatment modalities (18 meta-analyses), included 442 primary studies and 34,625. For differences between mental disorders and HCs, evidence was suggestive for 7 (13.2%) pooled analyses, indicating decreased HRV in dementia, PTSD, somatic symptom disorders, functional somatic syndromes, and schizophrenia. For other disorders, including autism spectrum disorder, alcohol use disorder, bipolar disorder, generalized anxiety disorder, insomnia, and major depressive disorder, the evidence was weaker and below the suggestive level. For treatment effects, 5 pooled analyses (27.8%) had weak evidence, indicating altered HRV before and after antipsychotic treatment, repetitive transcranial magnetic stimulation treatment, physiotherapy, and psychotherapy. The credibility of HRV evidence in mental disorders varied across HRV variables and diseases. No two diseases exhibited identical altered HRV patterns, highlighting the potential significance of overall HRV profiles in delineating distinct disorders.