Stretched diameter of the atrial septal defect (ASD), measured by balloon sizing, is generally used as a guide to the selection of the size of the device utilized for transcatheter closure of the ASD. Balloon sizing is a cumbersome procedure and sometimes requires the use of very large size balloon catheters. Several methods of assessment of ASD size, namely, echographic, pulmonary-to-systemic flow ratio (Qp:Qs), and angiographic measures, were undertaken in a group of 16 patients, aged 7 months to 45 years (median, 4.5 years), who were being evaluated for transcatheter closure of ASD; the results were compared with the stretched diameter. Although the echographic size of the ASD (9.9 +/- 4.1 mm, mean +/- SD) is similar (p greater than 0.1) to the angiographic size (7.9 +/- 2.5 mm), it is much smaller (p less than 0.01) than the stretched diameter (16.1 +/- 5.3 mm). When the relationship between various measures of ASD was examined, although the Qp:Qs ratio and angiographic size have a significant (p less than 0.05) correlation with the stretched diameter (r = 0.55 and 0.54, respectively), the echo diameter has the best correlation coefficient, r = 0.82, p less than 0.001. The stretched diameter can be estimated by the equation: 1.05 x echo + 5.49 mm. It is concluded that the echographic diameter is a useful adjunct in the estimation of the stretch ASD diameter, which in turn can be used in the selection of the size of the device for transcatheter occlusion of the ASD.