To investigate time-related association between fluid balance and prognosis in sepsis patients. A retrospective cohort study was conducted based on the data of sepsis patients in the Medical Information Database for Intensive Care-IV 2.0 (MIMIC-IV 2.0) from 2008 to 2019. Sepsis patients aged ≥ 18 years who were admitted to intensive care unit (ICU) for at least 2 days were included. The daily fluid balance and cumulative fluid balance (CFB) were calculated from days 1 to 7 after ICU admission. According to CFB,the patients were divided into negative fluid balance group (CFB% < 0%), fluid balance group (0% ≤ CFB% ≤ 10%), and fluid overload group (CFB% > 10%). In-hospital mortality was the primary outcome. Multifactorial Logistic regression was used to analyze time-related association between different CFB and the risk of in-hospital mortality in patients with sepsis during 7 days after ICU admission. In addition, subgroup analysis was performed on patients with septic shock and patients with sepsis who stayed in the ICU for 7 days or longer. A total of 11 437 patients with sepsis were included, of which 6 595 were male and 4 842 were female. The mean age was (64.4±16.4) years. A total of 10 253 patients (89.6%) survived and 1 184 patients (10.4%) died during hospitalization. Compared with the survival group, patients in the death group were older, lighter, had higher sequential organ failure assessment (SOFA), simplified acute physiology score II (SAPS II), longer ICU stay, higher incidence of septic shock, and higher proportion of invasive mechanical ventilation, renal replacement therapy (RRT) and vasoactive drugs. In terms of comorbidities, congestive heart failure, renal disease, liver disease, and malignancy were more common in the death group. The death group had a higher daily fluid balance than the survival group during 7 days after ICU admission, the CFB in the two groups gradually increased with length of ICU stay. After adjusting variables such as age, gender, race, SOFA score, SAPS II score, comorbidities, and the use of invasive mechanical ventilation, RRT and vasoactive drugs, multivariate Logistic regression analysis showed that fluid overload on day 1 after ICU admission was a protective factor for the reduced risk of in-hospital mortality in sepsis patients [odds ratio (OR) = 0.74, 95% confidence interval (95%CI) was 0.64-0.86, P = 0.001]. However, fluid overload on day 3 was a risk factor for in-hospital mortality in sepsis patients (OR = 1.70, 95%CI was 1.47-1.97, P < 0.001) and the risk of in-hospital mortality was significantly increased from day 4 to day 7. Furthermore, the same results were obtained in patients with septic shock and sepsis patients who stayed in the ICU for 7 days or longer. Fluid overload on day 1 was associated with reduced in-hospital mortality. However, from the third day, fluid overload increases the risk of in-hospital mortality. Thus, managing fluid balance at different times may improve prognosis.