To evaluate the influence of the size of the defect and the age of surgical repair on left ventricular mechanics, including geometry, shape, diastolic and systolic function as well as myocardial contractility, we used cross-sectional echo-Doppler to study 20 patients (12 males, 8 females) who had undergone successful surgical closure of a ventricular septal defect. The patients were divided in two groups, corrected early and late, on the basis of the degree of left-to-right shunting (ratio of pulmonary to systemic output of greater or less than 2.5/1) and the age at the surgical repair (older or younger than 2 years of age). The group undergoing early correction included 11 patients, mean age 7.1 ± 1.8 years (range 4.2–11.8 years) having surgery at mean age of 1.3±0.6 years for a large ventricular septal defect (mean ratio of pulmonary to systemic output of 3.1/1; range 3.4–2.7/1) with a mean postoperative follow-up 4.6±1.9 years. The group of nine patients undergoing late correction had a mean age of 11.3±4.9 years (range 6.7–17.2 years), with a later surgical repair (mean age 4.7±2.7 years) for a moderate-sized ventricular septal defect (mean pulmonary/systemic output ratio 2.1/1; range 2.3–1.7) and a mean postoperative follow-up of 7±4.2 years. Each group of surgically repaired patients was compared with a control group matched for age, body surface area and gender. No significant differences were found between the normal controls and those undergoing early correction for any assessed functional index regarding left ventricular geometry (normalized volumes and mass for body surface area, mass/volume and thickness/radius ratios), shape (long axis–short axis ratio), diastolic (mitral and pulmonary venous flow patterns) and systolic (fractional shortening and rate-corrected mean velocity of circumferential fibre shortening) function. In addition, the data points for each patient for the rate-corrected mean velocity of circumferential fibre shortening to end-systolic stress relationship were within the 95% confidence limits of normal, suggesting normal left ventricular contractility. On the other hand, the patients undergoing surgery at a later age showed a persistent increase of the normalized left ventricular end-diastolic volume and mass, with an higher mass/volume ratio and reduced end-systolic stress compared with normal controls. Furthermore, left ventricular shape (long axis–short axis ratio) was abnormal at end-diastole but with its normal values at end-systole. Our data suggest that, in the presence of a large ventricular septal defect, early successful surgical repair <2 years of age results in complete recovery of left ventricular mechanics in the postoperative follow-up. In ntrast, surgical closure at >2 years of age, even for a moderately sized ventricular septal defect, deleteriously affects postoperative left ventricular geometry and shape. Since prolonged volume overload may be detrimental to myocardial function, earlier surgical repair should be recommended.