The conduction time over the lower common pathway (LCP) in AVNRT can be assessed by subtracting the H A-interval during tachycardia (HAt) from that during ventricular pacing at exactly the same cycle length (HAp) (delta HA = HAp-HAt). It has been suggested that H-A measurements may help in the differentiation of Slow/Fast from Slow/Slow AVNRT. This study evaluated prospectively in 61 consecutive patients with AVNRT (43 +/- 15 y; 46 women, all with antegrade conduction during AVNRT over the slow pathway) how often a reliable measurement of the length of the LCP could be made, and in how far the results were concordant with mapping criteria for the differentiation of Slow/Fast from Slow/Slow AVNRT. A new para-Hisian pacing technique (using only the His bundle catheter) was applied in all patients. Comparison of HAt and HAp was possible in 44 of the 61 patients (72%). In these 44 patients, HAp was longer than HAt in 12 patients, indicating the presence of a LCP. All patients with delta HA > or = 15 ms had earliest retrograde atrial activation in the posterior septum (Slow/Slow AVNRT; n = 6) or simultaneously in the anterior and posterior septum (n = 1). On the other hand, 31 of the 32 patients without evidence of a substantial LCP (delta HA < or = 0) had typical Slow/Fast AVNRT. Moreover, although it appears logical for Slow/Fast AVNRT to have a shorter HAt than Slow/Slow AVNRT, an HAp of > or = 70 ms was a better discriminator between the two forms of AVNRT than any HAt value. Therefore, delta HA > or = 15 ms (sens. > or = 86%; spec. > or = 97%) or HAp > or = 70 ms (sens. = 100%; spec. > or = 89%) were highly indicative for the Slow/Slow variant of AVNRT. Using a para-Hisian pacing technique, H-A measurements can be performed in 72% of AVNRT patients. They can be used as an important tool in the differentiation of Slow/Fast and Slow/Slow AVNRT.