Objectives: Low total CO2 (tCO2) levels are significantly associated with all-cause mortality. Lots of factors are related to the poor prognosis of COVID-19, it was a lack of data to evaluate the impact of tCO2. We evaluated the impact of metabolic acidosis on all-cause mortality in patients with COVID-19. Methods: We retrospectively reviewed the data of in-hospital admitted patients with COVID-19 between December 2020 and August 2021. We excluded subjects with underlying end-stage kidney disease, no data of tCO2 value, and age under 18 years old. The primary outcome was in-hospital mortality. We evaluate the impact of tCO2 as a continuous variable on mortality using the Cox-proportional hazard model. In addition, we evaluate the relative factors associated with tCO2 <22 mmHg using Logistic regression analysis. Results: A total of 814 patients were included, and the mean age was 59.7±20.3 years old. Mean tCO2 was 26.2±3.6 mEq/L, and there were 75 (9.2%) with tCO2 <22 mEq/L. Increased in 1 mEq/L of tCO2 significantly decreased risk for all-cause mortality after adjustment with age, sex, history of hypertension, diabetes, and laboratory results such as serum white blood count (WBC), hemoglobin, platelet, glucose, albumin, eGFR, CRP (adjusted HR 0.95, 95% CI 0.90, 0.99). This result was prominent in age over 60, eGFR <60, and patients with hypertension in the subgroup analysis. Higher WBC, lower platelet, higher eGFR, lower serum sodium, and higher serum chloride are significantly associated with tCO2 <22 mEq/L. Conclusions: Decreased tCO2 significantly increased the risk of all-cause mortality in patients with COVID-19. Monitoring of tCO2 could be a good indicator to predict prognosis, and it needs to be considered to encourage in patients with a specific condition.