The response to adenosine 5'-triphosphate (ATP) identifies patients with syncope who might benefit from pacemaker therapy (ATP test). Two measures have been used to determine the outcome of the ATP test, which have lead to contrasting conclusions regarding its utility: (1) the duration of cardiac pause (CP) mainly due to AV block and (2) the longest RR interval (RRmax). We tested the hypothesis that the discrepancy regarding the utility of the ATP test is mainly because of the different way the 2 measures determine the outcome of the test. Post hoc analysis was applied to data obtained from patients with syncope (n = 33) with a positive and negative ATP test based on the CP duration and RRmax, respectively, subjected to pacemaker therapy. In 19 and 14 patients, the pacemaker was programmed to function as AAI pacing at 30 ppm (control) and as DDD pacing at 70 ppm, respectively. During the follow-up period of 17.0 ± 8.6 months, syncope recurred in only 1 of the 14 patients with DDD pacing; in contrast, 10 of 19 patients with AAI30 pacing experienced syncope within the first 5.3 ± 5.2 months of follow-up (P < 0.009; recurrence rate). The ATP test, the outcome of which is determined by the CP measure, is a useful diagnostic test for the identification of patients with bradycardic syncope who may benefit from pacemaker therapy; the identification of such patients would be missed when the RRmax measure is used to determine the outcome of the test. The efficacy of DDD pacing suggests that atrioventricular nodal conduction block is the primary cause of syncope in patients with a positive ATP test based on the CP measure.