We investigated 500 consecutive, unselected electrocardiograms of outpatients for interatrial block (TAB) using all 12 leads rather than the usual recommendation in the literature, which is lead IT, sometimes with another lead. TAB had been reported in 2 widely separated large general hospitals in > 40% of 1,000 patients in sinus rhythm in each. Because the P waves in TAB (duration >= 110 ms) generally have low amplitude despite their excessive width, we used magnifying graticules and, for greater specificity, a minimal duration of >= 120 ms. Four hundred sixty-nine patients remained after excluding those with atrial arrythmias or technically poor tracing. Two hundred three of these patients (40.6%) had TAB. Had we used lead IT alone, only 110 cases would have been identified, which would have meant overlooking almost 1/2 the cases with this lesion, which is important (1) as a predictor of atrial fibrillation and other arrhythmias, and (2) represents a large, dysfunctional left atrium. Leads V-3 and V-4 yielded larger numbers of TAB than lead II. (The slightly smaller prevalence than in the 2 cited studies may be due to our using 1/2 the number of patients.) Electrocardiographic interpreters should seek TAB in all 12 leads and onsider its anatomic functional and predictive correlates. (c) 2005 Elsevier Inc. All rights reserved.