To compare the effect of standard trauma craniectomy (STC) versus limited craniectomy (LC) on the outcome of severe traumatic brain injury (TBI) with refractory intracranial hypertension, we conducted a study at five medical centers of 486 patients with severe TBI (Glasgow Coma Scale score </= 8) and refractory intracranial hypertension. In all 486 cases, refractory intracranial hypertension, caused by unilateral massive frontotemporoparietal contusion, intracerebral/subdural hematoma, and brain edema, was confirmed on a CT scan. The patients were randomly divided into two groups, one of which underwent STC (n = 241) with a unilateral frontotemporoparietal bone flap (12 x 15 cm), and the second of which underwent LC (n = 245) with a routine temporoparietal bone flap (6 x 8 cm). At 6-month follow-up, 96 patients (39.8%) in the STC group had a favorable outcome on the basis of the Glasgow Outcome Scale, including 62 patients who had a good recovery and 34 who showed moderate deficits. Another 145 patients (60.2%) in the STC group had an unfavorable outcome, including 73 with severe deficits, nine with persistent vegetative status, and 63 who died. By comparison, only 70 patients (28.6%) in the LC group had a favorable outcome, including 41 who had a good recovery and 29 who had moderate deficits. Another 175 patients (71.4%) in the LC group had an unfavorable outcome, including 82 with severe deficits, seven with persistent vegetative status, and 86 who died (p < 0.05). In addition to these findings, the incidence of delayed intracranial hematoma, incisional hernia, and CSF fistula was lower in the STC group than in the LC group (p < 0.05), although the incidence of acute encephalomyelocele, traumatic seizure, and intracranial infection was not significantly different in the two groups (p > 0.05). The results of the study indicate that STC significantly improves outcome in severe TBI with refractory intracranial hypertension resulting from unilateral frontotemporoparietal contusion with or without intracerebral or subdural hematoma. This suggests that STC, rather than LC, be recommended for such patients.