Background. Weight loss in chronic heart failure is linked to impaired survival. We aimed to assess the frequency of weight loss in patients with this disease, whether the degree of weight loss predicts mortality, and whether weight loss can be prevented by angiotensin-converting-enzyme (ACE) inhibitors. Methods. We investigated weight changes in 1929 patients from the SOLVD trial who had chronic heart failure, were free of oedema at baseline, and survived for at least 4 months after trial entry. Mean follow-up was 35 months (SD 13). We analysed the effect of weight loss at cutpoints of 5%, 7·5%, 10%, 15% (a priori), and 6% (post hoc) to identify which one best predicted outcome. To validate results, we analysed data for 619 patients in the V-HeFT II trial. Findings. 817 (42%) patients in the SOLVD trial had weight loss from baseline of 5% or more. At 8 months follow-up, all cutpoints for weight loss were significantly associated with impaired survival after adjustment for age, sex, New York Heart Association class, left ventricular ejection fraction, and treatment allocation. Weight loss of 6% or more at any time during follow-up was the strongest predictor of impaired survival (adjusted hazard ratio 2·10, 95% CI 1·77-2·49; p<0·0001). Patients on the ACE inhibitor enalapril had a lower hazard of 6% or more weight loss than did those not taking the drug (adjusted reduction 19%, p=0·0054). Results from analyses of V-HeFT II data lent support to our findings. Interpretation. Weight loss occurs frequently in patients with chronic heart disease, its reversal is rare, and when present, it is independently linked to impaired survival. Weight loss of more than 6% should be used to define the presence of cachexia in patients with chronic heart failure. In chronic heart failure, treatment with an ACE inhibitor reduces the risk of weight loss.