We describe a novel methodological approach for inducing cold lesion in the mouse as a model of human cortical contusion trauma. To validate its reproducibility and reliability, dexamethasone (Dxm) was repeatedly applied to demonstrate possible antioedematous drug effects. Following the induction of anaesthesia with halothane, the dura was exposed via trephination. Using a micromanipulator a pre-cooled (-78 degrees C) copper cylinder, 3 mm in diameter, was pressed down to a depth of 1 mm onto the dura for 30 s under microscopic control. The body temperature was held constant at 37 degrees C throughout the procedure. Blood pressure (BP), measured by a modified photosensor-monitored tail-cuff method, and acid-base status were not significantly different when analysed before and after cold lesion and prior to sacrifice. However, there was a marginal mixed respiratory and metabolic acidosis. The antioedematous action of Dxm was studied in four standard pre-and post-treatment paradigms: 2x0.5 mg/kg (II), 2x12.5 mg/kg (III) and 4x6.25 mg/kg (IV: 3x pre-, 1x post-treatment: V: 1x pre-, 3x post-treatment). Physiological saline injections served as controls. High doses of Dxm (III-V) significantly attenuated the cold-lesion-induced loss of body mass. Dxm treatment also resulted in a reduction of brain water content (III; P<0.05), and brain swelling (IV; P<0.05) in the lesioned hemisphere, relative to controls. In conclusion, we have characterized a novel cold lesion model in the mouse to mimic traumatic brain injury and the beneficial effect of Dxm treatment on the extent of brain oedema.