This was analytic observasional study with retrospective cohort design. Data collection was done in Dr Soetomo Hospital?s isolation rooms during July-December 2020 period. Serial NLR and procalcitonin were recorded first on hospital admission, on highest value of NLR/procalcitonin, and then on 24 hours before hospital discharge. Secondary infection was recorded from positive cultures during treatment. Severity was measured by SOFA score and WHO. Mortality data were recorded from medical record. Spearman test, Mann-Whitney test, and ROC curve analysis were used in statistical analysis. There were 119 patients included in this study. Mortality rate was 38,6%. Secondary infection rate was 21,8. There were significant correlations between NLR and severity (p<0,001). NLR cutoff value as severe-critical illness predictors was >6,82. There were significant NLR differences between mortality groups (p<0,001). NLR cutoff value as mortality predictors was >12,75. There were significant correlations between procalcitonin and severity (p<0,001). Procalcitonin cutoff value as severe-critical illness predictors was >0,145 ng/ml. There were significant procalcitonin differences between mortality groups (p<0,001). Procalcitonin cutoff values as mortality predictors were >0,285 ng/ml. There were significant NLR and procalcitonin difference in patients with secondary infection. NLR and procalcitonin are useful as severity and mortality predictors in COVID-19.