The relative distributions of ventilation as measured by 99Tc-sulfur colloid aerosol deposition and pulmonary perfusion (measured with radiolabeled microspheres) were determined in 12 preterm lambs that were delivered at 138 days gestational age and ventilated for 4 hrs. To verify that unventilated lung segments in these lambs would have decreased perfusion, a balloon catheter was placed in a major bronchus either at birth or after 2 hrs of ventilation. This catheter prevented ventilation of 24.5 +/- 3.2% of the lung tissue. After 4 hrs of ventilation, the lambs were sacrificed and the lungs were divided into about 60 1-g pieces. Apart from the occluded, atelectatic segments, the lungs were visually well aerated with only 5.3 +/- 1.8% of the nonobstructed lungs being spontaneously atelectatic. There was a 66.5 +/- 0.07% decrease in blood flow to the area of lung made atelectatic by the balloon. The blood flow also was decreased to lung regions assessed to be spontaneously atelectatic. No 99Tc-sulfur colloid was recovered from balloon-occluded lung regions, and less 99Tc-sulfur colloid was found in the spontaneously atelectatic areas than in aerated lung regions. There were significant correlations (P less than 0.001) between pulmonary blood flow and aerosol recovery in each of the 12 animals. Premature lambs had a wide variability in ventilation and perfusion, but the relative ventilation to perfusion ratio was regulated to minimize the intrapulmonary shunt.