Adenosine can reveal dormant pulmonary vein (PV) conduction after PV isolation (PVI) in patients with paroxysmal atrial fibrillation (AF). However, the impact of elimination of adenosine-provoked dormant PV conduction after PVI has not been formally evaluated. The purpose of this study was to determine whether ablation of PV reconnections unmasked by adenosine improves outcomes. Patients with paroxysmal AF (n = 129) were randomized to receive either adenosine (n = 61) or no adenosine (n = 68) after PVI. Dormant conduction revealed by adenosine after PVI was ablated until all adenosine-mediated reconnections were eliminated. Thereafter, both groups received isoproterenol. Acute reconnection was seen in 23 patients (37%) in the adenosine group. There was a significant difference between the number of PVs reconnected if patients were given adenosine >60 minutes after initial PVI compared to those who received adenosine <60 minutes after initial PVI (3/32 [9.4%] vs 24/32 [75%], P <.0001). Patients who did not receive adenosine had more PV reconnections after isoproterenol infusion compared to patients in the adenosine group (17/68 [25.0%] vs 5/61 [8.2%], P = .018). There was no difference in the rate of AF recurrence in patients who received adenosine (24/61 [39%]) compared to control patients (23/68 [34%], log-rank P = .83). Adenosine can reveal dormant conduction in more than one-third of patients with paroxysmal AF undergoing PVI. However, adenosine administration, and additional ablation of the resultant connections, does not improve long-term outcomes compared to a protocol that includes isoproterenol infusion.