Near-infrared spectroscopy (NIRS) has been used to study cerebral haemodynamics and oxygenation in the preterm infant for many years, but its use as a clinical tool has remained elusive. This has partly been due to the challenges of providing a continuous quantitative measurement that is valid and reliable, as well as demonstrating that interventions based on NIRS measurements improve clinical outcome. Recent studies investigating cerebral oxygenation targeted treatment, and defining optimal blood pressure based on an assessment of cerebrovascular reactivity, suggest ways in which this technology may yet be clinically useful.