The efficacy of beta-blockers was tested in 11 patients who avoided sudden death as a result of ventricular tachycardia or fibrillation due to coronary artery disease or non-ischaemic underlying heart disease, after implantation of an automatic defibrillator. Ten patients initially tolerated acebutolol despite prior class III or IV heart failure in six cases; nadolol replaced acebutolol in nine cases for long-term therapy. In these 10 patients, periods of treatment with and without beta-blocking agent were available, making possible a crossover comparison, during a follow-up lasting 31.6 +/- 17.8 months. One hundred and ten shocks were delivered: 14 were considered as probably inappropriate and ruled out, leaving the remaining 96 shocks to be analysed. The monthly rate of shocks was lower during beta-blocking treatment: 0.12 +/- 0.24 vs. 1.09 +/- 1.41 (P = 0.03). While taking beta-blockers, only four patients received shocks, compared to 10 (i.e. all cases) not administered beta-blockers (P less than 0.01). Despite the technical limitations of the study, since only a few spontaneous shocks could be documented on ECG recordings, the efficacy of beta-blockers in preventing occurrence of severe ventricular tachyarrhythmias seems likely, and deserves further investigations using new implanted devices with improved memory functions.