The inert gas elimination technique was used to estimate pulmonary ventilation-perfusion (VA/Q) mismatching in heparinized, ventilated, anesthetized dogs during a 90-min period of hemorrhagic hypotension (mean arterial pressure 40 Torr) and subsequent reinfusion of the shed blood. Systemic and pulmonary arterial pressures, as well as cardiac output, were similar to those in previously reported studies using this protocol. Mean arterial O2 partial pressure (PO2) fell from 86 to 75 Torr after hemorrhage and rose to a mean value of 78 Torr after reinfusion. The VA/Q distributions showed that a mean of 56.7% of the ventilation was associated with unperfused or poorly perfused (VA/Q greater than 10) regions during hypotension (control 33.7%). After reinfusion, a mean of 47.8% of the ventilation was still directed to lung with little or no perfusion. This could not be explained on a hydrostatic basis, since pulmonary arterial pressure after reinfusion was greater than the control value. Shunt or blood flow to low VA/Q regions did not increase at any time during hemorrhagic hypotension or reinfusion. Microscopic examination of lung tissue revealed extensive leukocyte aggregation that was not seen in control animals. The mean diameter of obstructed pulmonary vessels was 35 microns (range 13.8-59.8 microns). Storing the shed blood in acid-citrate-dextrose instead of heparin had no significant effect on the extent of leukocyte aggregation. We suggest that leukocyte aggregation and margination may be related to the high VA/Q regions seen in these animals.