The relevance of timely detection and differential diagnosis of depressive disorder and bipolar affective disorder (BD) is beyond doubt. In complex differential diagnostic cases, clinicians can come to the aid of objective research methods, such as electroencephalography. This review summarizes and systematizes modern methods of differential diagnosis of unipolar and bipolar depression based on electroencephalography data. A literature search was conducted using the search terms «depression», «bipolar disorder», «unipolar depression», «differential diagnosis», «neuromarkers», «electroencephalography», «quantitative EEG». The search was conducted in the databases PubMed, NCBI, ResearchGate, RSCI, Scopus and Google Scholar in the period 2000 to 2022. In general, EEG indicators at rest in patients are characterized by ambiguity and require further clarification, however, there are some important findings on the differences in the asymmetry of electric activity, spectral power and coherence between patients with BD and unipolar depression. As for event-related potentials (ERP), similar results were obtained in a number of studies. There are a longer latency of P300 and a decrease in its amplitude in patients with BD compared to unipolar depression, which makes the ERP method more promising in solving the problem of differential diagnosis.