BACKGROUND: We analyzed data from COVID-19 patients in Japan to assess the utility of the 4C mortality score as compared with conventional scorings. METHODS: In this multicenter study, COVID-19 patients hospitalized between March 2020 and June 2021, over 16 years old, were recruited. The superiority for correctly predicting mortality and severity by applying the receiver operating characteristic (ROC) curve was compared. A Cox regression model was used to compare the length of hospitalization for each risk group of 4C mortality score. RESULTS: Among 206 patients, 21 patients died. The area under the curve (AUC) (95% confidential interval (CI)) of the ROC curve for mortality and severity, respectively, of 4C mortality scores (0.84 (95% CI 0.76-0.92) and 0.85 (95% CI 0.80-0.91)) were higher than those of qSOFA (0.66 (95% CI 0.53-0.78) and 0.67 (95% CI 0.59-0.75)), SOFA (0.70 (95% CI 0.55-0.84) and 0.81 (95% CI 0.74-0.89)), A-DROP (0.78 (95% CI 0.69-0.88) and 0.81 (95% CI 0.74-0.88)), and CURB-65 (0.82 (95% CI 0.74-0.90) and 0.82 (95% CI 0.76-0.88)). For length of hospitalization among survivors, the intermediate- and high- or very high-risk groups had significantly lower hazard ratios, i.e., 0.48 (95% CI 0.30-0.76)) and 0.23 (95% CI 0.13-0.43) for discharge. CONCLUSIONS: The 4C mortality score is better for estimating mortality and severity in COVID-19 Japanese patients than other scoring systems.