INTRODUCTION: Dual atrioventricular (AV) nodal physiology, defined as an AH jump > or =50 msec with a 10 msec decrease in A1A2, is the substrate for atrioventricular nodal reentrant tachycardia (AVNRT) and yet it is present in a minority of pediatric patients with AVNRT. Our objective was to characterize dual AV nodal physiology as it pertains to a pediatric population. METHODS/RESULTS: We retrospectively reviewed invasive electrophysiology studies in 92 patients with AVNRT (age12.1 +/- 3.7 yrs) and in 46 controls without AVNRT (age 13.3 +/- 3.7 yrs). Diagnoses in controls: syncope (N = 31), palpitations (N = 6), atrial flutter (N = 3), history of atrial tachycardia with no inducible arrhythmia (N = 3), and ventricular tachycardia (N = 3). General anesthesia was used in 49% of AVNRT and 52% of controls, P = 0.86. There were no differences in PR, AH, HV, or AV block cycle length. With A1A2 atrial stimulation, AVNRT patients had a significantly longer maximum AH achieved (324 +/- 104 msec vs 255 +/- 67 msec, P = 0.001), and a shorter AVNERP (276 +/- 49 msec vs 313 +/- 68 msec P = 0.0005). An AH jump > or =50 msec was found in 42% of AVNRT versus 30% of controls (P = 0.2). Using a ROC graph we found that an AH jump of any size is a poor predictor of AVNRT. With atrial overdrive pacing, PR > or = RR was seen more commonly in AVNRT versus controls, (55/91(60%) vs 6/46 (13%) P = 0.000). CONCLUSIONS: Neither the common definition of dual AV nodes or redefining an AH jump as some value <50 msec are reliable methods to define dual AV nodes or to predict AVNRT in pediatric patients. PR > or = RR is a relatively good predictor of AVNRT.