A number of studies have found an association between the increased concentration of glial fibrillar acid protein (GFAP) in blood serum in patients with various types of brain damage (ischemic stroke, traumatic brain injury, neurodegenerative and neuro-oncological diseases), as well as with a rapid decline in cognitive functions in elderly people with initially normal cognitive abilities.The objective: to identify the relationship between delayed cognitive recovery and changes in serum GFAP concentration in the perioperative period in patients operated for various oncological diseases.Subjects and Methods. The study included 30 patients who underwent surgical treatment for prostate cancer, colorectal cancer and pancreatic cancer under combined general anesthesia.The inclusion criteria were the expected duration of the operation over 300 minutes and the age over 60 years. GFAP was determined in plasma by enzyme immunoassay before anesthesia, the next day after surgery and on the 4th–5th day. Neuropsychological testing was performed before surgery and on the 4th–5th postoperative day. Delayed cognitive recovery was defined as a decrease in the composite z-score of more than one standard deviation (SD) compared to the preoperative assessment.Correlation analysis was performed between changes in the composite z-score (in absolute values) and the difference in GFAP concentration between the outcome and the first postoperative day, the outcome and the 4th–5th postoperative day and the first and 4th–5th postoperative days.Results. In five cases (16.6%), a decrease in the composite z-score > 1 SD was revealed indicating a delayed cognitive recovery. In the remaining 25 (83.4%) patients, changes in the composite z-score were less than one standard deviation. The median concentration of GFAP in patients with delayed cognitive recovery was 0.13 [0.1; 0.14] before surgery, 0.12 [0.09; 0.14] the day after surgery and 0.16 [0.05; 0.19] on the 4th–5th day after surgery. In patients without cognitive impairment, the concentration of GFAP was 0.15 [0.125; 0.184] before surgery, 0.15 [0.121; 0.163] 24 hours after surgery and 0.13 [0.079; 0.151] on the 4th–5th day after surgery. The correlation values between changes in the composite z-score and the difference in GFAP concentrations were: between the outcome and the first postoperative day – rs = 0.107, p = 0.37, outcome and the 4th–5th postoperative day – rs = 0.134, p = 0.37, the first and 4th‒5th postoperative days – rs = 0.21, p = 0.37.Discussion. There was no statistically significant difference in GFAP levels between patients with delayed cognitive recovery and patients without cognitive impairment. There was also no correlation between the difference in GFAP concentrations in plasma before surgery and 24 hours after, before surgery and on the 4th–5th day of the postoperative period and the composite z-score.Conclusions. The use of GFAP to predict cognitive decline associated with surgical treatment of colorectal cancer, prostate cancer and pancreatic cancer under general anesthesia is not yet possible.