The virtual electrode model predicts that pacing stimulus strength should reflect proximity of the pacing electrode to excitable myocardium, allowing pacing threshold to assess radiofrequency (RF) ablation lesions and unexcitable scar. The purpose of this study is to correlate RF lesion size with pacing threshold and electrogram (EG) amplitude change at the ablation site. In four swine (32-58 kg, 20 ventricular RF lesions were created using a 4-mm tip electrode catheters under fluoroscopic and electroanatomic guidance. Unipolar pacing threshold and bipolar and unipolar EG amplitude were measured before and after ablation and compared with lesion size measured in the fixed, serially sectioned tissue. Lesion diameter ranged from 6.4 to 19 mm and volume ranged from 29 to 1920 mm3. Ablation increased the pacing threshold by 320%, from 0.9 +/- 0.3 to 3.6 +/- 2.6 mA, P < 0.001. The change in pacing threshold correlated with lesion volume R = 0.88, P < 0.001). Linear regression predicts that lesion volume (mm3) = 160 X rise in pacing threshold + 13. Ablation reduced peak to peak bipolar EG amplitude by 56%, from 2.5 +/- 2.0 mV to 1.1 +/- 0.6 mV (P = 0.005). Unipolar EG amplitude diminished by only 22% from 4.0 +/- 1.6 to 3.2 +/- 0.9 mV postablation (P = 0.005). The correlations of lesion volume with change in either bipolar R = 0.14, P = 0.6) or unipolar R = 0.18, P = 0.6) EG amplitude were poor. Pacing threshold correlates with RF ablation lesion size, consistent with the virtual electrode model. In normal myocardium, change in pacing threshold is likely to be a better marker of lesion size than electrogram amplitude.