The measurement of ocular nerve sheath diameter (ONSD) via ocular ultrasound scanning is a recent non-invasive method for intracranial pressure (ICP) assessment. Few clinical studies have assessed ONSD variations during osmotherapy for the treatment of sustained increased ICP episodes. The aim of our study was to determine the rate of ONSD variation after mannitol administration for increased ICP episodes. We consecutively included in a prospective, observational study, the patients who had severe acute brain injury and monitored with an invasive ICP monitor. For each episode of sustained elevated ICP, the ONSD was measured in the right and left eye with a 7.5-MHz echography probe, and the mean value was reported. Simultaneously, ICP and cerebral perfusion pressure (CPP) were recorded. All measurements were performed just before and 20 minutes following a 20% mannitol infusion. Data were expressed as medians and interquartile ranges. Thirteen patients were included and analysed (traumatic brain injury, n=10; subarachnoid haemorrhage, n=3). The median value of the mannitol dose infused was 0.54 g/kg (0.49-0.80 g/kg). In all cases, the ONSD was greater than 5.8 mm before osmotherapy. The ONSD significantly decreased after mannitol infusion from 6.3 (6.1-6.7) to 5. mm (5.5-6.3) (p=0.0007). Concomitantly, the intracranial pressure decreased from 35 (32-41) to 25 (22-29) mmHg (p=0.001) and the CPP increased from 47 (50-60) to 66 (59-69) mmHg (p=0.003). The variations of ONSD appear to be an interesting parameter to evaluate the efficacy of osmotherapy for elevated ICP episodes in patients with acute brain injury.