Abstract Background Coronavirus disease (COVID-19) is a novel infectious disease developed into a pandemic since early 2020, resulting in extensive morbidity and mortality globally. Therefore, it is important to identify biomarkers for predicting the severity and mortality of COVID-19. We focused on presepsin (P-SEP) and thrombomodulin (TM), which are biomarkers of sepsis and endothelial dysfunction, respectively. Some reports have already suggested that these biomarkers can be associated with COVID-19, however, most of them assessed blood P-SEP and TM levels using only a single day measurement. In addition, since both P-SEP and TM are excreted by the kidney, renal dysfunction should be considered when evaluating these levels. In this study, to evaluate the usefulness of P-SEP and TM for the prognosis and prediction of COVID-19 severity, we investigated the changes in these levels after admission with or without adjustment for creatinine (CRE) level in COVID-19 patients. Methods Samples were residual sera obtained from COVID-19 patients (n = 674) admitted to the TMDU Hospital. Serum P-SEP and TM levels were measured using the STACIA clinical assay system. Disease severity was classified as mild, moderate, or severe based on hemoglobin oxygen saturation and the history of intensive care unit transfer or use of ventilation at admission. Patients in the severe group were further classified into survivors and non-survivors. This study was approved by the institutional research ethics committee (No. M2020-110). Results Since patients who received recombinant human TM treatment had abnormally high TM levels (> 1000 U/mL), data after administration were excluded from the analysis. P-SEP and TM levels during hospitalization were significantly correlated with CRE levels. In the correlation between P-SEP and CRE, the slope of the severe group was different from that of the non-severe group. When compared the data extracted within 5 days after admission (n = 183) among the three severity groups, CRE levels showed no significant difference among them. On the other hand, P-SEP and TM levels at admission were significantly higher in the severe group than those in the mild group, even after adjusting for CRE levels. In addition, there was no significant difference in the P-SEP levels between survivors and non-survivors, whereas TM levels were significantly higher in non-survivors than in survivors. Changes in the P-SEP levels at two time points (interval: 4.1 ± 2.2 days) were significantly different between survivors and non-survivors. The second measurement of TM levels were elevated in both groups from the first measurement at admission. Conclusion This study suggested that the elevation of P-SEP and TM levels in severe COVID-19 patients might not only fluctuate with renal dysfunction, but might reflect severity. TM and continuous P-SEP measurements are available for predicting mortality in patients with COVID-19. Moreover, P-SEP and TM values after CRE adjustment would be independent predictive markers, apart from renal function.