Background: Hyperosmolar therapy is widely used to treat traumatic intracranial hypertension, with mannitol as the mainstay of therapy. Its adverse effects on hemodynamics and renal function, as well as the "rebound̊ rise in intracranial pressure (ICP) observed after withdrawal or with prolonged use make it a less-than-optimal agent. Hypertonic saline has recently been introduced as a new hyperosmolar treatment for intracranial hypertension and cerebral edema in critical care units. Methods: The data for this descriptive study was collected prospectively. Subjects were entered as part of a compassionate use protocol after failing conventional therapy and barbiturate coma for control of elevated intracranial pressure. Parental consent was obtained for all subjects. Ten children (3 months to 12 years) with seven head injuries (Glasgow Coma Score 3-7) were entered into the study. Inclusion criteria included persistently elevated ICP > 20mmHg despite aggressive treatment tiling head elevation, paralysis and hyperventilation (pCo2 30-35mmHg), mannitol (serum osmolarities 315-320 mOsm/l), external ventricular drainage (4patients), and thiopental sodium. A continuous infusion of 3% saline on a "sliding scale" was used to achieve a target serum sodium level, which was adjusted to maintain ICP less, than 20 mmHg once the conventional therapy and barbiturate coma as outlined above failed to control intracranial hypertension. Serum sodium, osmolarity, ICP, MAP, CPP, and pCO2 were recorded as 6-hour intervals. If the patient had external ventricular drainage (4 patients), total daily CSF drainage was recorded. The total number of ICP "spikes" (>20 mmHg) were recorded for each 6 hour interval. Data was analyzed using chi square analysis and multivariate linear regression. Results: A significant decrease in ICP from pretrial baseline (mean 24 mmHg ± 5) was observed at 12 hours (mean 15 mmHg ±5) (p < 0.01) and 24 hours (mean11 mmHg ±6) (p < 0.01). A significant decrease in the number of ICP spikes per 6-hour period from pretrial baseline (mean 3 ICP spikes per 6 hour) was observed by Trial Day #2 (mean 1 ICP spike per 6 hour) (p < 0.01). Average protocol duration was 16 days. Mean serum sodium was 162 mEq/l ± 7 (range 140-187 mEq / l) and mean serum osmolarity was 347 mOsm/l ± 24 (range 315-407 mMol / l). Renal failure developed in two patients during episodes of sepsis and systemic inflammatory response syndrome (SIRS); both patients recovered full renal function before discharge from the hospital. Mean serum creatinine in eight patients who did not develop renal failure was 0.63±0.23 (mean±SD). Nine patients (90%) survived with average Glasgow outcome score of 4. Conclusion: Hypertonic saline appears to be promising adjunctive therapy to the treatment of traumatic intracranial hypertension. Hypernatremia and hyperosmolarity are safely tolerated in brain injured pediatric patients. Controlled trials are needed prior to recommendation of widespread use.