A growing body of evidence indicates there is an increasing incidence of cognitive dysfunction in patients after coronavirus disease 2019 (COVID-19) infection. However, still lack diagnostic tools, which allow us to predict prognosis in such cases and improve the stratification of the disease. This study aims to evaluate the usefulness of the biomarkers that could allow to predict the severity and progression of COVID-19 in patients with post-COVID syndrome and cognitive problems. Data regarding clinical history, pre-existing conditions, chest CT scan, and therapy (remdesivir, steroids) were acquired. A total of 44 patients with hospitalized COVID-19, and healthy controls were enrolled in the investigation, and serum blood was obtained. After 6 months of observations, patients with COVID-19 were divided into two groups: first - without post-COVID syndrome and memory complaints, and second - with post-COVID and cognitive problems. Measurements of YKL-40 and MR-pro-ADM were taken in the serum with enzyme immunoassay kits at the time of admission (visit 1) and 6 months after discharge from the hospital (visit 2). Significantly higher concentrations of YKL-40 were found in patients with COVID-19 as compared to healthy individuals (p = 0.016). Moreover, YKL-40 ratio allowed to differentiate patients with and without post-COVID syndrome (median: 0.94 vs. 1.55, p = 0.004). Additionally, COVID-19 patients with dyspnea presented significantly elevated levels of MR-pro-ADM as compared to the group of COVID-19 survivors without dyspnea (p = 0.015). In the group of patients without post-COVID syndrome, the concentrations of YKL-40 and MR-pro-ADM decreased after treatment as compared to levels before therapy (77 vs. 36 ng/ml and 607 vs. 456 pmol/L). However, in patients with post-COVID syndrome and cognitive problems, the levels of both markers did not alter 6 months after hospital discharge in comparison to basal levels. Furthermore, after dexamethasone treatment the YKL-40 concentrations declined significantly (p = 0.003) in patients with COVID-19. This study demonstrated the predictive usefulness of YKL-40 as an indicator of successful treatment in patients with COVID-19 infection allowing risk stratification of hospitalized patients. It seems that indicators of neuroinflammation might have the potential to track development of cognitive complaints, however, it requires further investigations.