The use of aggressive treatments and the modification of current treatment in patients with heart failure (HF) relies heavily on the assessment of disease severity using prognostic markers. However, many such markers are unavailable in routine clinical practice, and others have little prognostic value. This study tested the hypothesis that low body temperature could predict short-term survival after discharge in patients hospitalized for HF. Data from the Acute and Chronic Therapeutic Impact of a Vasopressin Antagonist in Congestive Heart Failure (ACTIV in CHF) trial, which randomized 319 patients hospitalized for HF to receive placebo or tolvaptan, were retrospectively analyzed. Hypothermia was defined a priori as an oral body temperature <35.8 degrees C at randomization. Cox regression was used to analyze survival within a 60-day follow-up period. Hypothermia was observed in 32 patients (10%). Mortality rates at 60 days after discharge were 6.3% (20 of 319) overall, 9.4% (3 of 32) in hypothermic patients, and 5.9% (17 of 287) in nonhypothermic patients. Hypothermia was a strong multivariate predictor of mortality; hypothermic patients were 3.9 times more likely to die within 60 days than nonhypothermic patients (95% confidence interval 1.002 to 15.16, p = 0.0497) after adjustment for treatment group, age, and other confounders. Hypothermia was associated with such indicators of low cardiac output as an elevated blood urea nitrogen/creatinine ratio, narrow pulse pressure, and a reduced ejection fraction. In conclusion, hypothermia appears to be a strong predictor of mortality in patients with HF.