SUMMARYWe conducted this study to verify the efficacy of atrial unipolar potentialfor for ablation of atrial tachycardia. The morphology of atrial unipolar potential at the successful and the best unsuccessful ablation sites was analyzed in 35 patients with atrial tachycardia (sino-atrial reentrant tachycardia (SART) 15, adenosine-sensitive atrial reentrant tachycardia near the His bundle (HAT) 10, and non-reentrant ectopic atrial tachycardia (EAT) 10). The usefulness of atrial unipolar potential was compared with the Ao-Po interval. The incidences of QS pattern at the successful and the best unsuccessful sites were (successful versus unsuccessful; P, respectively) 93 versus 55%; P = 0.20 in SART, 90 versus 0%; P = 0.0001 in HAT, and 90 versus 10%; P = 0.001 in EAT. The mean Ao-Po intervals at the successful and the best unsuccessful sites were 35 versus 30 ms; NS in SART, 48 versus 45 ms; NS in HAT, and 58 versus 50 ms; NS in EAT. A significantly higher incidence of transient success at QS pattern sites was observed in SART than in HAT or EAT (n = 2 ± 0.8 versus 0.2 ± 0.1 and 0.1 ± 0.1, P = 0.0005), and each transient site and final success site in SART was distributed linearly. The sensitivity and specificity of QS patterns with regard to the determination of appropriate target sites were 0.91 and 0.45 in SART, 0.9 and 1.0 in HAT, and 0.88 and 0.88 in EAT, respectively. In ablation of HAT and EAT, the QS pattern is very useful and should be given high priority when determining the optimum target site. In SART, the ablation success was often achieved by multiple, linear RF delivery near perinodal tissue, and the QS pattern could be a candidate for the optimum target site. (Int Heart J 2007; 48: 313-322)