Angiotensin converting enzyme (ACE) inhibitors are associated with a greater reduction in mortality in non-ischaemic cardiomyopathy than in ischaemic cardiomyopathy after the results of the V-HeFT-II and SOLVD trials in symptomatic patients. However, a recent analysis of the global, symptomatic and therapeutic, results of the SOLVD trials, demonstrated a similar reduction in mortality with ACE inhibitors in ischaemic and non-ischaemic cardiomyopathies. Moreover, after myocardial infarction, the beneficial effects of ACE inhibitors have been well established in patients with left ventricular dysfunction. Betablockers, especially bisoprolol in the CIBIS-I trial, also seem to be more effective in non-ischaemic cardiomyopathy. However, CIBIS-II and the US Carvedilol Heart Failure Trial Program clearly showed that the benefits of betablockade were identical whether ischaemic or not. The beneficial effects of betablockers in the post-infarction period are more marked when left ventricular dysfunction is severe. The PROVED and RADIANCE trials suggest that digitalis is more effective in non-ischaemic cardiomyopathy. These results were not confirmed by the DIG trial which showed a significant reduction in the combined criterion, mortality and hospital admission for aggravation of cardiac failure, both in ischaemic and in non-ischaemic cardiomyopathy. However, the use of digitalis should be prudent during ischaemic cardiomyopathy, the neutral effect on global mortality in the DIG trial masking divergent results with a tendency to reducing mortality due to aggravation of cardiac failure and a significant increase of other causes of cardiac death, especially from myocardial infarction and arrhythmias. Amiodarone could also be useful in non-ischaemic cardiomyopathy. The reduction in risk of death in the GESICA study, which comprised 60% of patients with non-ischaemic cardiomyopathy, contrasting with the absence of an effect with this molecule in the STAT-CHF trial which only comprised 29% of patients with non-ischaemic cardiomyopathy. The new generation of calcium antagonists could also be more effective in non-ischaemic cardiomyopathy. Although amlodipine significantly reduced mortality in the PRAISE trial in non-ischaemic cardiomyopathy, there was no favourable effect with felodipine in the V-HeFT-III tria. Finally, if in the earlier studies oral anticoagulants were more effective in non-ischaemic cardiomyopathy, the recent results of the SOLVD trial showed that warfarin decreased the mortality in both ischaemic and non-ischaemic cardiomyopathy. The value of anti-aggregant therapy is not questioned in coronary artery disease, but its role in dilated cardiomyopathy has not yet been established. In conclusion, apart from the use of digitalis which must be prudent in post-infarction cardiomyopathy or in patients with ventricular arrhythmias, the treatment of cardiac failure differs little with respect to its ischaemic or non-ischaemic aetiology, and should be based on the NYHA (New York Heart Association) classification.