The aim of this study was to assess if right atrial overdrive pacing can suppress symptomatic episodes of paroxysmal atrial fibrillation (PAF) in patients without bradyarrhythmias. Forty-two patients with frequent and symptomatic PAF without other pacing indication had a pacemaker implanted after a 4-week run-in period, during which the frequency of symptomatic PAF episodes and the mean heart rate were objectively documented. Depending on the mean heart rate recorded during run-in, the pacemaker was programmed in random order to right atrial AAI pacing at 10-19 beats/min > mean heart rate (medium overdrive [MO]), at 20-29 beats/min > mean heart rate (high overdrive [HO]) and to no pacing (OAO mode) for 4-12 weeks each using a crossover design. In the 35 patients who completed the protocol, the number of symptomatic episodes of PAF (>30-second duration) per week was significantly lower during MO pacing (median 0.88, P = 0.001, n = 35) and during HO pacing (median 0.75, P = 0.002, n = 20) than during OAO (median 2.02 and 2.04, respectively). There was no difference between MO and HO pacing in the 20 patients paced at both rates (0.97 vs 0.75, P = 0.33). Seven patients did not complete the protocol due to persistent atrial fibrillation (n = 4), angina pectoris requiring surgery (n = 1), and unwillingness to continue the study due to improvement (n = 1) or worsening (n = 1) of symptoms during the study periods. Right atrial overdrive pacing can reduce the number of symptomatic PAF episodes in patients with frequent and drug refractory PAF but without bradyarrhythmias.