Background: To determine disease severity in the COVID-19 pandemic, it is necessary to monitor several clinical and laboratory parameters. Non-invasive and objective inflammatory markers can be a practical and objective route in the determination of disease activity. The C-reactive protein/albumin ratio (CAR) is an inflammatory marker thought to have prognostic value for most diseases, primarily sepsis. The aim of this study was to evaluate the diagnostic performance of CAR in determining the clinical severity of COVID-19 disease. Methods: We conducted a retrospective examination of health records of patients diagnosed with COVID-19 between 15 March and 26 May 2020. Patients with pulmonary computed tomography findings consistent with COVID-19 and confirmed by a positive result in the real-time reverse transcriptase-polymerase chain reaction test were included. The cases were separated into two groups of mild and severe according to symptom severity. Results: Of the 105 patients, 39 (37.1%) were males with a mean age of 63.2±14.5 years. In the regression analysis, the severity of COVID-19 was associated with age (OR = 1.04; 95% CI: 1.001, 1.074; p = 0.04), heart disease (OR = 4.37; 95% CI: 1.02, 18.8; p= 0.03), CAR (OR = 1.77; 95% CI: 1.05, 2.98; p = 0.03), lymphocyte count (OR = 0.54; 95% CI: 0.32, 0.96; p = 0.02). For the diagnosis of severe disease, the highest sensitivity of CAR was 76.5%, specificity 76.1% and the greatest area under the curve (AUC) was 0.70. Conclusion: We found a significant relationship between disease severity and CAR in COVID-19 patients. CAR is an inexpensive and practical marker for the diagnosis of severe COVID-19.