Atrial and ventricular pacemaker function was studied in 20 patients with idiopathic chronic complete heart block using the rate response to an intravenous bolus dose of isoprenaline (5 mug/70 kg bodyweight). Pacemaker responses were compared with those of 16 normal control subjects. None of the patients was having syncopal attacks at the time of admission and they were therefore selected in that none required immediate pacing. Ten of the patients had His bundle electrograms; all were shown to have a pre-His type of atrioventricular block. Two major groups emerge from the responses to isoprenaline. (a) High risk group: 11 of the 14 patients with reduced ventricular pacemaker responses had frequent syncopal attacks; 8 of the patients with Adams-Stokes syncope had a bundle-branch block pattern, while 3 had a narrow QRS. These patients require pacing. (b) Low risk group: a low risk asymptomatic group (5 patients) was identified with atrial and ventricular responses to isoprenaline within normal range. One of these patients had a bundle-branch block pattern, while 4 had a narrow QRS. These patients might be managed without pacing. The atrial response to isoprenaline was reduced in 12 of the 20 cases, 10 of whom also had reduced ventricular responses. All 9 patients with bundle-branch block had reduced ventricular responses, while 7 had reduced atrial responses. This evidence indicates that cardiac conducting tissue pathophysiology is widespread in complete heart bolck. The present work suggests that consideration of the ventricular pacemaker function is important in assessing liability to syncope in complete heart block. While patients with Adams-Stokes attacks require pacing it is suggested that all asymptomatic patients with complete heart block and those with minor symptoms are assessed using studies of both ventricular pacemaker function and site. A low risk group not requiring a pacemaker may emerge after sufficient follow-up assessment.