Background. There is no specific biomarker used in the diagnosis of COVID-19 and predicting its clinical severity. This study aimed to investigate the utility of ischemia-modified albumin (IMA) in diagnosing and predicting clinical severity in children with COVID-19. Methods. Between October 2020 and March 2021, 41 cases constituted the COVID-19 group and 41 cases constituted the healthy control group. IMA levels were measured at admission (IMA-1) and 48-72 hours (IMA-2) in the COVID-19 group. In the control group, it was measured at admission. COVID-19 clinical severity was classified as asymptomatic infection, mild, moderate, severe, or critical disease. Patients were divided into two groups (asymptomatic/mild and moderate/severe) to evaluate IMA levels in terms of clinical severity. Results. In the COVID-19 group, the mean IMA-1 level was 0.901±0.099, and the mean IMA-2 level was 0.866±0.090. The mean level of IMA-1 in the control group was 0.787±0.051. When IMA-1 levels of COVID-19 and control cases were compared, the difference was statistically significant (p<0.001). When clinical severity and laboratory data are compared, C-reactive protein, ferritin and ischemia-modified albumin ratio (IMAR) were statistically significantly higher in moderate-severe clinical cases (p=0.034, p=0.034, p=0.037 respectively). However, IMA-1 and IMA-2 levels were similar between the groups (p=0.134, p=0.922, respectively). Conclusions. To date, no study has been conducted on IMA levels in children with COVID-19. The IMA level may be a new marker for the diagnosis of COVID-19 in children. Studies with a larger number of cases are needed to predict clinical severity.