Aim To describe the significance of serum albumin for chronic heart failure and its prognosis. Methods We enrolled 820 consecutive patients from cardiology department of Shengjing hospital, and from whom venous blood for liver function test measurements and other examinations were drawn immediately upon hospital admission. 391 cases were CHF patients, and the other 429 cases were non-CHF patients as the control group. Then a prospective study was performed, considering 391 patients with chronic heart failure. We used multivariable modelling to assess the relationships between baseline serum albumin and heart failure and its prognosis. During a mean follow-up time of 14.46 （1～42） months, our study showed that the incidence of poor prognosis （cardiovascular death or heart failure hospitalization） was 27.2%, with death occurring in 28 patients （7.6%）. The Kaplan-Meier, Log Rank and Cox tests were used. Results In univariate analysis, serum albumin concentration was significantly lower in the CHF group than that in non-CHF group（37.49±4.60 g/L vs 41.46±3.59 g/L, P<0.001）. A stepwise logistic regression analysis showed that decreased albumin concentration was a risk for CHF（OR0.859（0.789～0.935）,P<0.001）. In the prospective study and univariate analysis, decreased albumin was a strong predictor of poor prognosis for both the composite outcome of cardiovascular death and heart failure hospitalization（HR0.941（0.903～0 .981）,P0.004）. Even after adjustment for other covariates, decreased albumin was one of the strongest independent predictors of poor prognosis （HR0.945（0.894～1.000）,P0.049）. Conclusion Serum albumin concentration was significantly lower in the CHF group than that in non-CHF group, and decreased albumin was a strong predictor of poor prognosis for both the composite outcome of cardiovascular death and heart failure hospitalization. Changes in serum albumin concentration may offer insight into the underlying pathophysiology of chronic heart failure.