Spirometric and blood gas studies have shown that pulmonary function is significantly impaired during the course of bronchography but improves rapidly once the patient is allowed to expectorate (1–3). The decrease in the measured lung volumes indicates that the impaired pulmonary function is due, at least in part, to obstruction of the bronchi with contrast material. In addition, it has been demonstrated experimentally that perfusion of a segment of lung will diminish if its alveolar oxygen is depleted (4–6). Also, in acute and perennial asthma, marked abnormalities in the distribution of pulmonary perfusion develop, as indicated by perfusion lung scans (7, 8). This is associated with, and probably secondary to, airway narrowing. Since experimental hypoxia as well as bronchial disease can disrupt the distribution of pulmonary perfusion, it appears reasonable that any material which accumulates in the bronchial tree impairing ventilation also can alter the distribution of perfusion. A perfusion lung scan, which was taken on the same day as a bronchogram and which revealed perfusion defects not seen in a study three months earlier, stimulated the present study. The radiograph obtained at the time of the second scan showed residual bronchographic contrast medium (Fig. 1). This raised the question whether the new perfusion defects represented pulmonary emboli or were due to the residual contrast material in the bronchi. Methods Nine patients who were scheduled for bronchography because of suspected bronchiectasis were studied (TABLE I). Eight had perfusion scans prior to bronchography. In 5 of the 8, repeat scans were obtained approximately one hour following bronchography. Two patients were studied the day after bronchography. In the eighth patient, bronchography was performed twice, four weeks apart. Perfusion scans were made before each bron-chogram, the day after the first bron-chogram, and one hour following the second. One patient also had a ventilation study prior to and one hour following bronchography. The ninth patient had ventilation and perfusion studies after bronchography only. The distribution of pulmonary ventilation was determined while the patients were in the sitting position. A normal tidal volume of air tagged with 2 m Ci xenon gas was inhaled, and an image was obtained with a gamma scintillation camera2 during brief breath-holding. Images were obtained in the posterior and one-side projections. Perfusion studies were performed with the intravenous injection of 1.5 to 2.0 m Ci 99mTc MAA (macroaggregated albumin) in the sitting position. The posterior and side views were obtained with the scintillation camera; these were followed by rectilinear photoscans. Radio-graphs were obtained simultaneously with each scan.