Intermittent atrial undersensing is observed in a considerable percentage of patients with single lead VDD pacemakers. Analyzing the 2-year data of the Saphir Multicenter Follow-Up Study, the authors investigated predictors for the occurrence of undersensing. The study included 194 patients with high degree AV block who received a VDD pacemaker system with an identical sensing amplifier. Placement strategy of the atrial dipole was left to the discretion of the implanting physician. At the final position, atrial potential amplitudes were measured during deep and shallow respiration. Atrial dipole position was determined by intraoperativefluoroscopy subdividing the right atrium in a high, mid, and low portion. Undersensing was defined by evidence of at least one not sensed P wave during Holter monitoring or exercise testing and by the presence of 0.1-0.2 mV amplitudes in the P wave amplitude histogram of the pacemaker. Incidence of undersensing was 25.8%; 9.3% of patients showed frequent (> 5%) or symptomatic undersensing. Patients with undersensing were older (76.6 +/- 10.6 vs 64.2 +/- 14.8 years), showed a lower minimum of intraoperative atrial potential amplitude (P(min) 0.86 +/- 0.64 vs 1.43 +/- 0.77 mV), a wider range of potential amplitude (deltaP 1.71 +/- 1.44 vs 0.94 +/- 0.84 mV), and a higher incidence of dipole placement in the low right atrium (50.0% vs 11.1 %, P < 0.001 for all comparisons). In a multivariate regression analysis, patient age > 66 years, Pmin < 0.6 mV, > 1.3 mV and atrial dipol placement in the lowright atrium were independently predictive for undersensing. Minimal atrialpotential amplitude, range of potential amplitude, and atrial dipole position influence atrial sensing performance in single lead VDD pacing. Thus, implantation guidelines should reflect these rules to improve the outcome of VDD pacemaker recipients.