Nine hundred eleven patients with systolic blood pressures less than or equal to 90 mm Hg were randomized to MAST and No-MAST groups, and all taken to a single Level I Trauma Center. Epidemiologic analysis of etiology, age, race, sex, Trauma Scores, and Injury Severity Scores revealed the two groups to be statistically identical. Seven hundred eighty-four patients were analyzed in detail. The principal injury location was thorax in 41%, abdomen in 32%, extremity in 16%, head in 7%, and neck in 4%. In patients with head and extremity wounds, the mortality rate was clearly not a function of MAST use. In the chest, abdomen, and neck, MAST did not improve survival. Although the total prehospital time was 35.8 + 10.4 minutes for MAST patients and 32.5 +/- 10.7 minutes for No-MAST patients, 70% of patients with MAST had prehospital times greater than 30 minutes compared to 55% of the No-MAST patients. When the prehospital time was greater than 30 minutes, 31% of the MAST patients died, compared to 21% of the No-MAST patients. MAST application adversely affected the outcome most significantly for patients with cardiac and thoracic vascular injury. The overall mortality of 31% in the MAST group, compared to 25% in the No-MAST group was statistically significant (p = 0.05).