Studies of experimental traumatic brain injury have found that histologic injury is significantly reduced and behavioral outcome improved in animals treated with moderate hypothermia (32-33 degrees C) or a 21-aminosteroid. Because these treatments are thought to work through different neurochemical mechanisms, their combined use might be expected to be more efficacious than either treatment alone. To test this hypothesis, we studied each treatment separately and in combination in a rodent model of controlled cortical contusion. Treatment with moderate hypothermia (32 degrees C for 4 h), a 21-aminosteroid (U-74389G, Upjohn, 10 mg/kg intravenously, repeated 3 h after the first dose), or both, was initiated 10, 25, or 40 min after injury. The brains were perfused 24 h after injury, and sagittal sections were stained for the 68-kDa "core" neurofilament subunit, a marker of axonal injury. The total number of positively stained axons in the ipsilateral internal capsule were counted under light microscopy. Compared with the control group (injury but no treatment), treatment with each therapy alone or combined, initiated 10 min after injury, caused significant reductions in the number of stained axons (21-aminosteroid alone-35% reduction; hypothermia alone-55% reduction; combination-48% reduction). The number of positively stained axons was significantly reduced in the aminosteroid and combination therapy groups at all three postinjury times (p = 0.01) and in the hypothermia groups treated at 10 or 25 min (p = 0.01) but not at 40 min after injury. We concluded that combination therapy with hypothermia and 21-aminosteroids was no more efficacious than either therapy alone, and that 21-aminosteroid therapy was more efficacious than hypothermia when treatment was initiated 40 min after injury.