Hypoalbuminemia is a long-term risk factor for incident of both myocardial infarction and heart failure. We assessed whether serum albumin levels at admission are associated with new-onset heart failure and in-hospital mortality in patients with acute coronary syndrome (ACS). The study included 7,192 patients with ACS with no previous history of heart failure. Patients were divided into quartiles according to serum albumin levels (Q1: ≤3.50 g/dl; Q2: 3.51 to 3.80 g/dl; Q3: 3.81 to 4.08 g/dl; and Q4: >4.08 g/dl). Logistic regressions were used to explore the relations among serum albumin quartiles, new-onset heart failure, and in-hospital mortality. Serum albumin levels were negatively correlated with both high-sensitivity C-reactive protein and white blood cell count at admission. The unadjusted rate for both new-onset heart failure (37.7%, 20.2%, 14.7%, and 11.4% for Q1, Q2, Q3, and Q4, respectively; p <0.0001) and in-hospital mortality (9.8%, 3.4%, 2.0%, and 1.7% for Q1, Q2, Q3, and Q4, respectively; p <0.0001) were higher at lower serum albumin levels. Multivariate analysis demonstrated that serum albumin level ≤3.50 g/dl is an important and independent predictor of both the development of new-onset heart failure (odds ratio 2.31, 95% CI 1.87 to 2.84, p <0.0001) and in-hospital mortality (odds ratio 1.88, 95% CI 1.23 to 2.86, p = 0.003). In conclusion, albumin level ≤3.50 g/dl is an independent predictor of new-onset heart failure and in-hospital mortality in patients with ACS. The inflammatory state may be a mechanism underlying hypoalbuminemia in this clinical setting.