The right ventricle (RV) is the main ventricular chamber in many congenital heart diseases before and after surgical correction, and it is the most important determinant of outcome in postoperative tetralogy of Fallot and other complex malformations. Unfortunately its irregular crescentic shape does not allow the use of the geometric assumption used for the left ventricle. Many methods have been suggested in the literature to overcome this problem, none fully reliable. The introduction of volume-rendered 3-dimensional (3D) reconstruction of echocardiography images provides a tool for the direct measurement of cardiac chambers, not based on geometric assumptions. The aim of this research study was to determine the accuracy of 3D echocardiography (3DE) to measure RV volumes in pediatric patients with secundum atrial septal defects, compared with direct volume measurements performed during the intervention. We performed 3DE study in the operating department, with the patient anesthetized, intubated, and ventilated before the surgical procedure. Sequential 2-dimensional echocardiographic images for subsequent 3D rendering were acquired using an ultrasound machine with a transthoracic 4-MHz rotational or 5-MHz transesophageal omniplane probe; in the last 5 patients a machine was used that was equipped with a 3600-crystal real-time 3D probe. To validate the 3DE measurements, these were compared with the volume of the RV directly measured in the operating department, at the end of the surgical procedure, injecting saline solution through the tricuspid valve, using a graduate syringe. Among 25 pediatric patients enrolled in the study, with an age range of 1 and 14 years (mean 4 years) and a weight range of 8.5 to 57.4 kg (mean 18.6 kg), in 23 a mean of 3 echocardiographic acquisitions were performed and compared with the direct measurement. A close comparison was found between RV volumes measured by 3DE and direct volume measurements (P < .00001). The regression line, shifted toward the y axis, which describes the 3DE volumes, indicated that the echocardiographic measures overestimate the surgical ones. In our study this overestimation had the mean of 9% with values comprised between 3% and 19%. The coefficient of repeatability was 4.79 mL with all the values within this range (2 SD of the mean). We conclude that 3DE provides an accurate measurement of RV volume in pediatric patients with RV volume overload. It is a reliable, noninvasive, and nongeometric method of evaluation of the volume of this chamber, and can be considered a precious tool in the armamentarium of the pediatric cardiologist.