Neurogenic pulmonary edema is a serious and always life-threatening complication following several lesions of the central nervous system. We report an experiment with 58 Wistar-Hanover adult male rats. Two groups were formed: control (n=4) and experimental (n=54). The experimental group sustained acute midthoracic spinal cord injury by Fogarty's balloon-compression technique containing 20 microL of saline for 5, 15, 30 or 60 seconds. The rats were anesthetized by intraperitoneal (i.p.) sodium pentobarbital (s.p.) 60 mg/Kg. The quantitative neurological outcome was presented at 4, 24 and 48 hours from compression to characterize the injury graduation in different groups. Poor outcome occurred with 60 seconds of compression. Six animals died suddenly with pulmonary edema. Using the procedure to investigate the pulmonary edema during 60 seconds of compression, followed by decompression and time-course of 60 seconds, 20 rats were randomly assigned to one of the following groups: control (1, n=4, anesthetized by i.p. s.p., 60 mg/Kg but without compression) and experimental (2, n=7, anesthetized by i.p. xylazine 10 mg/Kg and ketamine 75 mg/Kg) and (3, n=9, anesthetized by i.p. s.p., 60 mg/Kg). The pulmonary index (100 x wet lung weight/body weight) was 0.395 +/- 0.018 in control group, rose to 0.499 +/- 0.060 in group 2, and was 0.639 +/- 0.14 in group 3. Histologic examination of the spinal cord showed parenchymal ruptures and acute hemorrhage. Comparison of the pulmonary index with morphometric evaluation of edema fluid-filled alveoli by light microscopy showed that relevant intra-alveolar edema occurred only for index values above 0.55. The results suggest that the pulmonary edema induced by spinal compression is of neurogenic nature and that the type of anesthesia used might be important for the genesis of lung edema.