Background: Early fluid expansion could prevent postoperative organ hypoperfusion. However, excessive fluid resuscitation adversely influences multiple organ systems. This retrospective, observational study aimed to investigate the relationship between early negative fluid balance and postoperative mortality in critically ill adult patients following cardiovascular surgery. Methods: In total, 567 critically ill patients who had undergone cardiovascular surgery and whose intensive care unit length of stay (LOS) was more than 24 hours were enrolled. The baseline characteristics, daily fluid balance and cumulative fluid balance were obtained. Patients were followed until discharge or day 28. Multivariate logistic regressions adjusted by propensity score were used to analyze the relationship between early negative fluid balance and postoperative mortality. Results: Overall, postoperative mortality was 6.2% (35/567). Acute Physiology and Chronic Health Evaluation II on admission (odd ratios [OR] 1.110), acute kidney injury stage (OR 1.639) and renal replacement therapy received (OR 3.922) were the independent risk factors of postoperative mortality, whereas negative daily fluid balance at day 2 (OR 0.411) was the protective factor. Patients with a negative daily fluid balance at day 2 had lower postoperative mortality (3.4% vs. 12.2% in the positive fluid balance group), lower acute kidney injury (AKI) stage, were less likely to receive renal replacement therapy (RRT) and experienced shorter hospital LOS compared with those with a daily positive fluid balance. Conclusion: This retrospective, observational study indicates that early negative fluid balance is associated with lower postoperative mortality in critically ill patients following cardiovascular surgery. Further prospective, randomized trials are needed to prove the benefits from the restrictive fluid management strategy.