A questionnaire survey of 66 patients with funnel chest who underwent corrective surgical procedures by the sternal elevation method, with or without the application of metal strut, demonstrated that the operative result was good in 60.6% and fair in 39.4%. None of the patients rated the result as unsatisfactory. A computed tomogram of the chest wall was performed to study the depression (b/c), asymmetry (b'/b), and flatness (a/b) of the chest wall, where a was the maximum transverse distance of the chest wall, b and b' were the maximum distance from the anterior to the posterior chest wall at the left and right sides (b greater than b'), and c was the perpendicular distance from the point of the anterior chest wall at its greatest deformity to the level of the anterior tip of the spine. In patients with a good result, b/c and b'/b were well corrected, while in patients with a fair postoperative result, they were still significantly different from those in subjects with normal chest walls. Moreover, 85.7% of the patients (6/7) with b/c over 3.0 before operation had a fair postoperative result. The degree of a/b was not corrected in patients with either good or fair postoperative results. We conclude that an operative approach to lengthen ribs would be necessary to improve the degree of a/b, that in patients with severely depressed funnel chest, expressed as a b/c value over 3.0 by computed tomography, a transient support with struts should be applied, and finally, that a more careful approach for correction of asymmetry should be undertaken to improve the operative results.