Psychiatric conditions, such as mood and anxiety, are complex and heterogeneous syndromes that encompass varied, co-occurring symptoms and divergent responses to treatment. To facilitate characterizing the heterogeneity of neuropsychiatric disorders, we used a multimodal functional imaging approach to examine human brain connectivity. Here, we investigated data from the Tulsa-1000 study, including electroencephalographic (EEG) recordings, functional magnetic resonance imaging (fMRI) and structural MRI data, behavior tests, and psychological and clinical assessments acquired from 288 subjects. The electrical brain activity was analyzed using the temporal independent EEG microstates (EEG-ms), a technique that can measure dynamic neural activity. Using a data-driven approach, we examined the properties of EEG-ms in both Mood and Anxiety (MA) patients and healthy controls (HC) subjects. We found distinctive EEG-ms associated with the dorsal default mode network and anterior salience network in both MA group and healthy controls. Specifically, the occurrence rate of the dorsal default mode network was positively correlated with rumination response in the MA group. Our results reveal a novel finding of abnormal neural dynamics and contribute to the underlying pathophysiological mechanisms of MA. Implications of this technology can provide new insights into understanding the biological mechanism of mood and anxiety disorders.