A 57-year-old man with repetitive long R–P supraventricular tachycardia (SVT) was referred for radiofrequency catheter ablation. SVT was sensitive to adenosine triphosphate, was not linked between atrial and ventricular activation, and did not terminate with the ventricular response; thus, we speculated reentrant atrial tachycardia (AT). Although the mitral annulus (MA) was the earliest atrial activation site (EAAS; 5 o’clock), AT was ablated not at the EAAS, but at the tricuspid annulus (TA; 4 o’clock), 24 ms later than the EAAS. We suggest that preferential conduction from the TA to the EAAS of the MA may be involved in AT.