Pulmonary function tests were performed before surgery on 152 patients who were operated on for pectus excavatum between 1970 and 1987 and at long-term follow-up to assess the degree of impairment and to investigate any changes caused by surgical correction. The mean age at surgery was 15.3 +/- 5.5 years. Pulmonary function was found to be restricted preoperatively. Multivariate analysis showed that preoperative pulmonary function was not related to age, the severity of the deformity at physical examination, or to pulmonary complaints. Only the patients with obstructive disease showed significantly more pulmonary complaints (p = 0.042). The total lung capacity (TLC) and inspiratory vital capacity (IVC) were significantly related to the age-corrected (delta) anteroposterior diameter of the chest (lower vertebral index [LVI]) (p = 0.0001). At follow-up (mean, 8.1 +/- 3.6 years), the restriction of pulmonary function was increased despite improvement in the symptoms of most patients and despite a significant increase in the anteroposterior diameter of the chest (p = 0.0001): the TLC was decreased from 83.7 percent predicted (pred) preoperatively to 73.8 percent pred (p = 0.0001) and the IVC from 78.3 percent pred to 70.7 percent pred (p = 0.0001). The surgical results were satisfactory in 83.6 percent. No relation was found between the changes in pulmonary function measured at follow-up and the surgical results. Only the age at surgery and the changes in the TLC and IVC at follow-up were significantly related (p = 0.0036, 0.0043, respectively), although the correlation coefficients were low (r = 27 percent and 28 percent, respectively). The reduction in lung function at follow-up was most pronounced in the patients who had the least functional impairment (TLC > 75 percent pred) preoperatively. No correlation was found between the changes in the pulmonary function test results at follow-up and follow-up interval, preoperative delta LVI, and the change in delta LVI at follow-up.