Objective: Reliable staging of dementia is critical for both clinical practice and research. Although the Clinical Dementia Rating (CDR) is widely used to assess dementia severity, it is subject to subjectivity and inter-rater variability. Resting-state electroencephalography (EEG) may provide objective neurophysiological information associated with clinical severity. Methods: Resting-state EEG data from 600 adults aged ≥60 years in Korea were analyzed. Relative power spectral density was extracted from 19 electrodes across six frequency bands. Ordinal logistic regression and linear regression analyses exam-ined associations between resting-state EEG spectral features and CDR Global and CDR Sum of Boxes (CDR-SB) scores. All models adjusted for age, sex, education, and depressive symptoms, with false-discovery-rate correction. Results: Ordinal logistic regression demonstrated that higher resting-state theta-band EEG power was significantly associatedwith greater CDR Global severity, particularly in frontal, temporal, central, and parietal regions. Consistently, linear regressionanalyses revealed that higher CDR-SB scores were associated with increased theta in widespread cortical areas. No significantassociations were observed in other frequency bands. Conclusion EEG spectral slowing—marked by elevated theta power—was consistently related to CDR-defined dementia se-verity. Resting-state EEG can potentially serve as an objective neurophysiological correlate that informs clinician-rated staging and enhances the reliability of dementia severity assessment.