BACKGROUND Arrhythmogenic right ventricular dysplasia (ARVD) is characterized by recurrent ventricular tachycardia of right ventricular origin and a cardiomyopathy with hypokinetic areas involving the free wall of the right ventricle. Subjects have a risk of sudden cardiac death, particularly during sports and strenuous exercise. Routine clinical examinations may be normal, but fragmented or delayed electrograms are usually recorded in the right ventricle of these patients. However, the frequency with which late potentials are detected by conventional time-domain analysis of the signal-averaged ECG (SAECG) is not high. This study evaluated the usefulness of the frequency-domain analysis of the SAECG in addition to the conventional time-domain analysis for a screening test to detect patients with ARVD. RESULTS SAECG was recorded by using a bipolar X, Y, and Z lead system in 28 patients with ARVD (mean age, 38 +/- 13 years) and 35 age-matched normal subjects (mean age, 35 +/- 11 years). The conventional time-domain analysis of the SAECG was performed at two different high-pass filter settings, 25 and 40 Hz, and the low-pass cutoff frequency was fixed at 250 Hz. The fast-Fourier transform analysis of SAECG was performed using a Blackman-Harris window. Area ratio 1 (area of 20 to 50 Hz)/(area of 0 to 20 Hz) and area ratio 2 (area of 40 to 100 Hz)/(area of 0 to 40 Hz) were calculated. In the conventional time-domain analysis, 20 (71%) and 18 (64%) patients had positive criteria at filter settings of 25 and 40 Hz, respectively. In the frequency-domain analysis, 18 (64%) and 20 (71%) patients had abnormal values in area ratios 1 and 2, respectively. Combining the time- and frequency-domain analyses, all patients were judged positive, with a sensitivity of 100% and a specificity of 94%. CONCLUSIONS Each result of the time- and frequency-domain analyses revealed that both methods had equivalent value. Combining the two domain analyses improved the sensitivity without reducing the specificity. These findings suggest that combining the time- and frequency-domain analyses of the SAECG may be useful as a screening test to detect patients with ARVD.