Exercise training is currently recommended in the management of mild hypertension, but the relationship between training and ventricular arrhythmias has never been investigated in hypertensive subjects. Forty hypertensive sportsmen were studied by means of 24-h ECG Holter monitoring and the results were compared with those obtained in 40 sedentary hypertensives, 40 normotensive sportsmen and 40 normotensive sedentary subjects. Among the hypertensive sportsmen 82.5% exhibited at least one ventricular extrasystole and 32.5% complex forms of ectopy, a prevalence higher than that observed in the sedentary hypertensives (50% and 17.5%; P = 0.002). In the normotensive sportsmen the prevalence of ventricular arrhythmias (62.5% and 22.5%) was lower than that in the hypertensive sportsmen, but the difference was not statistically significant. During a training session the prevalence of ventricular ectopy was similar in the two groups of trained individuals. Among the hypertensive sportsmen no correlation was found between the severity of ventricular arrhythmias and the degree of left ventricular hypertrophy and performance. The results of the present study suggest that exercise training may enhance left ventricular vulnerability in hypertensive subjects. Whether subjects who manifest complex ventricular arrhythmias should continue to train remains a matter for individual judgement.