Experiments in 157 rats were carried out to establish that dose and time of initial dexamethasone therapy after head injury to counteract brain edema are decisive for the antiedematous effect (dose/time response relationship). For this purpose various human doses (weight-related conversion): 20 mg, 100 mg, 500 mg or 2500 mg of dexamethasone with different intervals from the injury: 10 min, 20 min, 40 min or 80 min, were administered in the model of cold lesion induced brain edema. Injury-induced (without therapy), the edema and sodium values increased markedly, and the potassium values decreased. Administration of dexamethasone produced statistically significant dose- and time-dependent effects tending to achieve physiological conditions: An optimal dose of 500 mg of dexamethasone had the highest antiedematous effect, while with still higher doses the effects must be expected to recede again or even be damaging. As expected, pretraumatic dexamethasone doses had the greatest antiedematous effect: reduction by 49%. On administration up to about half an hour after the injury, clearly relevant effects (up to 28%) were still measurable. With longer time intervals between injury and initial dexamethasone administration, e.g. 1 1/2 hours, a measurable but less relevant edema reduction by about 10% can be expected. All posttraumatic effects were achieved experimentally with a maximal dexamethasone therapy period of 21 hours. If similar results are obtained in corresponding clinical studies, the practical recommendation--at least from the animal experimental viewpoint--would be to administer ultra-high cortisone doses (e.g. 500 mg of dexamethasone) as early as possible within the first 2-3 hours after head injury. A cortisone therapy period of more than 2 or 3 days does not appear appropriate. In general no side effects are to be expected with this therapy regimen.