OBJECTIVES--To describe neurohormonal activation in patients with mild or moderate heart failure and how it may be modified by treatment with ramipril. SETTING--Cardiology departments at 24 hospitals in Denmark, Finland, Norway, and Sweden. PATIENTS--223 patients with mild or moderately severe congestive heart failure who were taking diuretics with or without digitalis. DESIGN--Randomised, double bind, multicentre, placebo controlled comparison of ramipril and placebo. Venous blood samples were drawn at rest, before blind treatment, and after 12 weeks of treatment with the study drug. A probability prediction score for mortality derived by stepwise linear discriminant from neurohormone data in the first cooperative north Scandinavian enalapril survival study (CONSENSUS I) was used to assess combined activity of the different neurohormonal systems. RESULTS--Plasma concentrations of atrial natriuretic peptide were raised at baseline but angiotensin II, aldosterone, and noradrenaline concentrations were within normal limits. There was, however, a wide interindividual variation. Plasma noradrenaline concentration and prediction score were higher among patients with class III congestive heart failure according to the New York Heart Association's classification than among patients with class II congestive heart failure (P < 0.05). There was a modest but significant inverse correlation between exercise duration at baseline and plasma noradrenaline concentration (r = -0.21, P = 0.0023), aldosterone concentration (r = -0.14, P = 0.04), and prediction score (r = -0.24, P = 0.0004). Prediction score at baseline was significantly higher among those who died (n = 10) than among survivors (P = 0.03). Angiotensin converting enzyme activity was suppressed and plasma concentrations of aldosterone and atrial natriuretic peptide were reduced after 12 weeks of treatment with ramipril compared with placebo. In patients with the most pronounced neurohormonal activation at baseline (highest third of noradrenaline concentration or prediction score), noradrenaline concentration and prediction score were significantly lower after 12 weeks of taking ramipril compared with placebo. Patients with a prediction score in the highest third at baseline had a higher heart rate than to those in the lowest third (P = 0.003). CONCLUSIONS--Neurohormonal activation is associated with the degree of symptoms and the severity of disease in mild or moderately severe congestive heart failure. Treatment with ramipril attenuates neurohormonal activation. This effect is most pronounced among patients with the highest circulating concentrations of neurohormones before the start of treatment.