The role of pulmonary innervation in the genesis of hemodynamic responses to hypertonic salt solutions was assessed in an animal model of total lung denervation by total division of the pulmonary hilum followed by reimplantation of the organ. This was performed in 10 mongrel dogs (weighing 12 to 20 kg) randomly assigned to two groups: group I (five dogs) was comprised of animals with catheters placed in the pulmonary artery of the denervated lung; group II (five dogs) was comprised of animals with catheters placed in the pulmonary artery of the intact lung; a control group (group III) (five dogs) was submitted to a sham thoracotomy with catheters inserted in either pulmonary artery. On the seventh postoperative day the mean arterial pressure (MAP) was monitored and severe hemorrhagic shock (MAP = 40 mm Hg) was produced in all animals. After 30 minutes of shock the shed blood was discarded and 5% of the shed volume (+/- 2 ml/kg) was infused through the pulmonary catheter in the form of a hypertonic NaCl solution (2400 mosm/L). MAP continued to be measured for the 30 minutes following the infusion period. A significant rise of MAP was uniformly observed in animals of groups II and III. In group I low elevations of MAP were observed during the infusion period, followed by a return to shock levels on discontinuation of the infusion. The results suggest that selective lung denervation abolished the beneficial cardiovascular effects of hypertonic NaCl infusion during resuscitation from severe hemorrhagic shock without affecting the plasma osmolality pattern.