BACKGROUND: Bipolar disorder (BD) and major depressive disorder (MDD) share overlapping depressive symptoms, which pose challenges for achieving rapid and accurate differential diagnosis in clinical practice. This study aims to investigate whether alpha sub-band activity in electroencephalogram (EEG) can serve as a discriminative feature between MDD and BD, thereby improving diagnostic accuracy in mood disorders. METHODS: This study recruited a total of 103 participants, comprising 37 patients diagnosed with MDD, 36 patients with BD, and 30 healthy controls (HC). All participants were matched in terms of gender and age. EEG data were acquired during both eyes-open and eyes-closed states over a 5-minute duration to examine whether different sub-oscillations in the alpha band can differentiate between MDD and BD. RESULTS: We found that at the group level, the peak frequency of the HC group was in the low alpha band, the BD group in the medium alpha band, and the MDD group in the high alpha band. Our results indicate that the MDD and BD groups display the most pronounced differences in the high alpha band, irrespective of whether the eyes are open or closed. In contrast, the HC group exhibits some distinctions from the MDD and BD groups in the low alpha band. CONCLUSIONS: This study provides novel insights into the differential characteristics of alpha sub-band oscillations in MDD from BD as compared to healthy controls. These observations suggest distinct neural signatures for MDD and BD, highlighting the potential value of alpha sub-band analyses in diagnostic classification.