The primary goal of this study was to determine the incidence of post-traumatic ventriculomegaly (Evans' index > or = 0.30) in 95 head-injured patients with a Glasgow Coma Scale (GCS) score of < or =13 at admission. Additional objectives were to determine the relationship between an increase in ventricular size and several clinical and radiological features and outcome. A planimetric study was carried out in the sequential control computed tomography (CT) scans of 34 moderately head-injured (GCS 9-13) and 61 severely head-injured (GCS 3-8) patients with a minimum follow-up of 2 months. Between two and six CT scans were evaluated in each patient. The presence of subarachnoid hemorrhage (SAH) was registered. Evans' index was determined in all CT scans. In the final CT scan of each patient, ventricular size was related to the admission GCS score, age, the presence of SAH in the initial CT scans, type of brain lesion (classified according to the final diagnosis in the Traumatic Coma Data Bank classification), and outcome. Ventriculomegaly was found in 39.3% of patients with severe head injury and in 27.3% of those with a moderate head injury. Increased ventricular size was evident 4 weeks after injury in 57.6% and 2 months after injury in 69.7%. No relationship was found between post-traumatic ventriculomegaly and age, initial GCS score, the presence of SAH, or type of lesion (focal or diffuse). Post-traumatic ventriculomegaly was significantly correlated with outcome. Post-traumatic ventriculomegaly is a frequent and early finding in patients with moderate or severe traumatic brain injury.