The aim of this study was to evaluate the clinical significance of ventricular tachycardia (VT) induced by atrial pacing. A group of 145 patients with spontaneous and induced VT was studied. Twenty-four VTs were induced by atrial stimulation (Group I) and 121 by ventricular stimulation (Group II). The underlying cardiac disease was comparable in the two groups (ejection fraction 32 +/- 14% versus 34 +/- 17%). Spontaneous bi-tachycardias, syncope and VT induced by exercise testing were more common in Group I. The prognosis was worse with 7 cardiac deaths in Group I versus 23 in Group II; recurrences of VT were commoner in group I. In this group, an electrophysiological (branch to branch or fascicular reentry) or clinical mechanism (ischaemia or poor haemodynamic status) could usually be demonstrated. Some cases of idiopathic VT were also observed in young patients. These results suggest that atrial stimulation should be performed routinely during electrophysiological studies of VT because the induction of VT by this method is not uncommon (16%). In addition, these cases of VT usually have a precise mechanism and a poor prognosis and, therefore, an appropriate treatment should be given.