In a prospective randomized double-blind multicenter trial, the efficacy and safety of a 51-hour ultra-high intravenous dexamethasone dosing regimen was investigated in patients with moderate and severe head injury. 300 patients between 15 and 55 years of age were randomized to receive either placebo or dexamethasone: 500 mg intravenous infusion within 3 h after trauma initially, followed by 200 mg after 3 h, thereafter 8 times 200 mg at 6 hourly intervals, resulting in a total administered dose of 2,3 g in 51 hours. Primary end points for assessment of efficacy were: Modified Glasgow Coma Scale (grading 3-16) on Day 5, modified Glasgow Outcome Scale (grading 1-6) 10-14 months after injury, and the time interval until consciousness improved above a level of modified GCS > or = 8. Secondary endpoints were CT results and neurological and laboratory data. The two groups were well matched with respect to important prognostic variables, such as age, severity of trauma, and interval between trauma and application of the drug. 269 patients (89.7%) were available for final examination after 10-14 months. Results were surprisingly favourable in both groups: Lethality in the dexamethasone and placebo group was 14.3 and 15.4%, respectively, and 61.7 and 57.4%, respectively, achieved social and professional rehabilitation after 10-14 months (outcome scale 6). No statistical difference was seen between the dexamethasone and the placebo group in any of the primary end points of efficacy and safety (incidence of upper gastrointestinal bleeding, infection, and thrombosis).(ABSTRACT TRUNCATED AT 250 WORDS)