As a component of treatment planning for thoracic irradiation (RT), 210 bronchogenic carcinoma patients seen at the Fox Chase Cancer Center from 1983 to 1990 underwent quantitative perfusion scans, superimposition of their RT treatment fields onto these scans, and pulmonary function testing. These studies were used to prospectively estimate the influence of the planned thoracic irradiation on pulmonary function, as measured by the forced expiratory volume in one second (FEV1). Among the 156 patients with unresected lesions, the mean pre-RT FEV1 was 1.71 +/- 0.67 liters (+/- standard deviation), and the mean percentage of total lung perfusion within the treatment field was 31.0 +/- 12.1%. Mean values for the 54 patients treated post-operatively were 1.79 liters (pre-RT FEV1) and 28.8% (% perfusion within RT field). Using this technique, the prospectively predicted post-RT FEV1 is the product of the pre-RT FEV1 (1% of total lung perfusion within the treatment field). The mean predicted post-treatment FEV1 for the nonoperative patients was 1.15 +/- 0.43 liters and 1.25 +/- 0.41 liters for the postoperative patients. Forty-three nonoperative and 19 postoperative patients had FEV1 determinations following RT, at a mean post-RT interval of 11 months for nonoperative patients and 23 months for post-operative patients. Among nonoperative patients, 53% had no change in post-RT FEV1, 19% improved, while 22% had readings declining toward the predicted value. Only 5% had readings below predicted. Among postoperative patients, 37% had no change or improvement, 37% declined toward the predicted, 10% declined to predicted, and 11% had values worse than predicted. This technique of superimposing RT fields onto lung perfusion scans predicts for a degree of pulmonary impairment which is observed in only a minority of patients (10%) and which is rarely exceeded (6%).