Eighty patients sustaining head injuries and presenting with Glasgow Coma Scale scores of 8 or less were entered into a prospective randomized study to assess the benefit of intracranial pressure (ICP) monitoring with two regimens of mannitol administration. Group I was treated with mannitol for ICP elevations greater than 25 mm Hg, while Group II received empirical mannitol therapy irrespective of ICP readings. No statistically significant differences in mortality rate or neurological outcome were demonstrated between the two groups. These results are comparable to those of several published series of head-injured patients receiving similar treatment from 1977 to 1982. However, those series must be reassessed in light of recently published studies with treatment initiated at lower levels of ICP.