Aim. To study the existence of clinical and humoral variables that behave as predictors of mortality risk in patients with dilated cardiomyopathy (DCM) and heart failure (HF) functional class III/IV New York Heart Association. Material and methods. We consecutively included 84 patients with DCM and HF over a year old of diagnosis (mean 3.3 years/SD 2.1), hospitalized at coronary care unit due to cardiac decompensation. Patients were divided into two groups: not dead group (GV) n=60, and group of those who died (GM) n=24. At the end of the study we compared: rehospitalizations, comorbidities, hygienic-dietary treatment accomplishment, adherence to pharmacological treatment, pharmacological treatment at income (drugs), and biochemical variables: natremia, white blood cell count, glucose, creatinine in blood, bilirubinemia. Results. Every patient had cardiothoracic index >0.5 and moderate to severe impairment of ventricular function by echocardiogram among both groups. Within sex distribution there was a masculine predominance (GV: 83.3% and GM: 70.8%; p=NS), age GV: 61.3 +/- 11.0 years old and GM: 63.9 +/- 11.3 (p=NS). Conclusion. Of all analyzed variables, were associated with worst prognosis: the largest number of readmissions, treatment with furosemide, leukocytosis (white blood cells > 11000 per mm(3)) and creatinine values > 2 mg/dL.