Mechanism-based sepsis treatments are unavailable, and their incidence is rising worldwide. Deaths occur during the early acute phase of hyperinflammation or subsequent postacute hypoinflammatory phase with sustained organ failure. The acute sepsis phase shifts rapidly, and multiple attempts to treat early excessive inflammation have uniformly failed. We reported in a sepsis cell model and human sepsis blood leukocytes that nuclear NAD+ sensor SIRT1 deacetylase remodels chromatin at specific gene sets to switch the acute-phase proinflammatory response to hypoinflammatory. Importantly, SIRT1 chromatin reprogramming is reversible, suggesting that inhibition of SIRT1 might reverse postacute-phase hypoinflammation. We tested this concept in septic mice, using the highly specific SIRT1 inhibitor EX-527, a small molecule that closes the NAD+ binding site of SIRT1. Strikingly, when administered 24 h after sepsis, all treated animals survived, whereas only 40% of untreated mice survived. EX-527 treatment reversed the inability of leukocytes to adhere at the small intestine MVI, reversed in vivo endotoxin tolerance, increased leukocyte accumulation in peritoneum, and improved peritoneal bacterial clearance. Mechanistically, the SIRT1 inhibitor restored repressed endothelial E-selectin and ICAM-1 expression and PSGL-1 expression on the neutrophils. Systemic benefits of EX-527 treatment included stabilized blood pressure, improved microvascular blood flow, and a shift toward proimmune macrophages in spleen and bone marrow. Our findings reveal that modifying the SIRT1 NAD+ axis may provide a novel way to treat sepsis in its hypoinflammatory phase.