INTRODUCTION: Bipolar voltage is used during electroanatomic mapping to define abnormal myocardium, but the effect of activation rate on bipolar voltage is not known. We hypothesized that bipolar voltage may change in response to activation rate. By examining corresponding unipole signals we sought to determine the mechanisms of such changes. METHODS AND RESULTS: LA extrastimulus mapping was performed during CS pacing in 10 patients undergoing 1(st) time paroxysmal atrial fibrillation ablation. Bipolar and unipolar electrograms were recorded using a PentaRay catheter (4-4-4 spacing) and indifferent IVC electrode, respectively. An S1S2 pacing protocol was delivered with extrastimulus coupling interval reducing from 350 ms to 200 ms. At each recording site (119±37 per LA), bipolar peak-to-peak voltage, unipolar peak to peak voltage and activation delay between unipole pairs was measured. Four patterns of bipolar voltage/extrastimulus coupling interval curves were seen: voltage attenuation with plateau voltage >1 mV (48±15%) or <1 mV (22±15%), and voltage unaffected by coupling interval with plateau voltage >1 mV (17±10%) or <1 mV (13±8%). Electrograms showing bipolar voltage attenuation were associated with significantly greater unipolar voltage attenuation at low (25±28 mV/s vs 9±11 mV/s) and high (23±29 mV/s vs 6±12 mV/s) plateau voltage sites (P<0.001). There was a small but significant increase in conduction delay between unipole pairs at sites showing bipolar voltage attenuation (P = 0.026). CONCLUSIONS: Bipolar electrogram voltage is dependent on activation rate at a significant proportion of sites. Changes in unipolar voltage and timing underlie these effects. These observations have important implications for use of voltage mapping to delineate abnormal atrial substrate. This article is protected by copyright. All rights reserved.