Abstract: Objective To investigate the EEG grading (EEG) of severe cerebrovascular disease (SCVD) and to analyze its relationship with clinical prognosis, with a view to provide a reference for the clinical prognosis of SCVD. Methods A total of 102 SCVD patients were selected from June, 2015 to March, 2017 and treated in neurology department of Traditional Chinese Medical Hospital of Zhejiang Province. EEG was monitored for these patients, including the Glasgow Coma Scale (GCS) and EEG Surveillance, to analyze the relationship between EEG classification and prognosis, GCS score and prognosis; the relationship between the related factors and the prognosis of patients with SCVD was studied by Logistic regression analysis. Results All 102 patients were followed up for 7 days and 48 patients survived and 54 patients died; EEG grade Ⅰ patients in 51 cases, 43 cases of survival, 8 cases of death, grade Ⅱ patients in 9 cases, 5 cases of survival, 4 cases of death, Ⅲ, Ⅳ, Ⅴ, Ⅵ grade patients all died, the mortality rate was 100%, EEG grading prognosis was 48 cases of survival, 54 cases died; Through the GCS score scale, found in 102 patients, the score was less than 3 points in 60 cases, GCS score was 2.63±0.32, the score was greater than 3 points in 42 cases, GCS score was 5.98±0.86; The area under the ROC curve (AUC) of EEG and GCS were 0.882 and 0.715, and the AUC values of the EEG and GCS were significantly different (P<0.05); SCVD prognosis was associated with GCS score and EEG grade. Conclusion EEG monitoring and grading results used in clinical can better reflect the severity of SCVD, and can better predict the prognosis, and it is worthy of clinical application.