Steroids are currently being used in sepsis, particularly in septic shock. However, clinical trials to date have shown contradictory results. This could be attributed to the different patient endotypes and steroid doses, which have also contributed to the inconclusive results. We investigated the effects of glucocorticoid therapy on sepsis in a polymicrobial sepsis model in a variety of settings, such as steroid dose, severity, and sepsis phase. We used a rat model of fecal slurry polymicrobial sepsis. First, we investigated the optimum dose of steroids in a sepsis model. We administered different doses of dexamethasone after sepsis induction (0.1DEX; 0.1 mg/kg, 0.2DEX; 0.2 mg/kg, 5DEX; 5 mg/kg). Second, we used two different severities of the fecal slurry polymicrobial sepsis rat model to examine the effects of the steroids. A moderate or severe model was defined as a survival rate of approximately 70% and 30%, respectively. Third, we administered steroids in an early (1 h after sepsis induction) or late phase (25 h after sepsis). In all the experiments, we investigated the survival rates. In the determined optimal model and settings, we measured serum lactate, alanine transferase (ALT), creatinine, tumor necrosis factor-α (TNF-α), interleukin (IL)-6, IL-10, and arterial blood gas. We evaluated the bacterial burden in the blood and spleen. Endotoxin tolerance of peripheral blood mononuclear cells (PBMCs) and splenocytes was also investigated to determine the level of immune suppression 24 h after sepsis by measuring TNF-α production after stimulation with lipopolysaccharide (LPS) in an ex vivo model. Early treatment of 0.2 mg/kg dexamethasone in a severe sepsis model showed the best beneficial effects. In moderate- or late-phase sepsis, there was no survival gain with steroid treatment. DEX 0.2 group showed less acute kidney injury manifested by serum creatinine and blood urea nitrogen. DEX decreased the levels of cytokines, including IL-6, IL-10, and TNF-α. Colony-forming units were significantly decreased in the blood when administered with dexamethasone. Endotoxin tolerance was not significantly different between the DEX 0.2 and control groups. In conclusion, early treatment of 0.2 mg/kg dexamethasone improved the outcomes of rats in a severe sepsis model.