Twenty-month-old male Fischer 344 rats with bilateral contusions of the frontal cortex (n=8/group) or sham operations received 16 mg/kg of progesterone or vehicle at 1 and 6 h post-injury, then once every 24 h for the next seven days, with tapering of the dose over the final two treatments. The rats' behavioral recovery was then evaluated on tests of locomotor activity and Morris water maze learning. All rats were sacrificed 21 days post-TBI and brains were perfused and cryoprotected for necrotic cavity measurements. The injury produced significant impairments in activity and spatial learning compared to sham controls. The progesterone-treated rats had better functional outcomes than vehicle-treated rats with similar cortical injuries. The neurosteroid treatment did not affect the size of the necrotic cavity.