To examine the spatial resolution of unipolar pace-mapping, 12 lead electrocardiograms (ECGs) recorded during pacing from each of the poles of a quadripolar catheter (5 mm interelectrode distance) were examined. Unipolar pacing was performed from each of the poles at late diastolic threshold, twice threshold and 10 mA at a cycle length of 500 ms. In 15 patients, pacing was performed at the right ventricular apex and in 14 at various left ventricular sites. Pacing from the distal catheter pole at threshold (index ECG) was used to simulate the site of origin of ventricular tachycardia, and all other ECGs were compared with the index ECG. Electrocardiograms were evaluated by two independent observers for 1) minor configuration differences (notch, new small component, change in the amplitude of individual components or change in QRS shape); 2) major differences in configuration (new large component, marked change in the amplitude of an existing component or two minor changes); and 3) peak to peak changes in amplitude. Minor differences in configuration were seen in a mean 2.4 +/- 1.9, 4.6 +/- 2.4 and 4.4 +/- 2.9 leads during pacing at 5, 10 and 15 mm from the distal electrode (index site). Major differences in configuration were seen in a mean of 0.3 +/- 0.5, 2.1 +/- 2.1 and 3.7 +/- 2.3 leads during pacing at 5, 10 and 15 mm from the index site. Differences in amplitude were seen in a mean of 3.1 +/- 2.2, 5.6 +/- 2.5 and 6.8 +/- 3.0 leads per ECG during pacing at 5, 10 and 15 mm from the index ECG pacing site, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)