Monitoring of intracranial pressure (ICP) is an important and integrated part of the treatment algorithm for children with severe traumatic brain injury (TBI). Guidelines often recommend ICP monitoring with a treatment threshold of 20 mmHg. This focused review discusses; (1) different ICP technologies and how ICP should be monitored in pediatric patients with severe TBI, (2) existing evidence behind guideline recommendations, and (3) how we could move forward to increase knowledge about normal ICP in children to support treatment decisions. Current reference values for normal ICP in adults lie between 7 and 15 mmHg. Recent studies conducted in "pseudonormal" adults, however, suggest a normal range below this level where ICP is highly dependent on body posture and decreases to negative values in sitting and standing position. Despite obvious physiological differences between children and adults, no age or body size related reference values exist for normal ICP in children. Recent guidelines for treatment of severe TBI in pediatric patients recommend ICP monitoring to guide treatment of intracranial hypertension. Decision on ICP monitoring modalities are based on local standards, the individual case, and the clinician's choice. The recommended treatment threshold is 20 mmHg for a duration of 5 min. Both prospective and retrospective observational studies applying different thresholds and treatment strategies for intracranial hypertension were included to support this recommendation. While some studies suggest improved outcome related to ICP monitoring (lower rate of mortality and severe disability), most studies identify high ICP as a marker of worse outcome. Only one study applied age-differentiated thresholds, but this study did not evaluate the effect of these different thresholds on outcome. The quality of evidence behind ICP monitoring and treatment thresholds in severe pediatric TBI is low and treatment can potentially be improved by knowledge about normal ICP from observational studies in healthy children and cohorts of pediatric "pseudonormal" patients expected to have normal ICP. Acceptable levels of ICP - and thus also treatment thresholds-probably vary with age, disease and whether the patient has intact cerebral autoregulation. Future treatment algorithms should reflect these differences and be more personalized and dynamic.