OBJECTIVES: Retrospective data from patients with severe traumatic brain injury (TBI) indicate that deviation from the continuously calculated pressure reactivity-based "optimal" cerebral perfusion pressure (CPPopt) is associated with worse patient outcome. The objective of this study was to assess the relationship between prospectively collected CPPopt data and patient outcome after TBI. METHODS: We prospectively collected intracranial pressure (ICP) monitoring data from 231 patients with severe TBI at Addenbrooke's Hospital, UK. Uncleaned arterial blood pressure and ICP signals were recording using ICM+® software on dedicated bedside computers. CPPopt was determined using an automatic curve fitting procedure of the relationship between pressure reactivity index (PRx) and CPP using a 4-h window, as previously described. The difference between an instantaneous CPP value and its corresponding CPPopt value was denoted every minute as ΔCPPopt. A negative ΔCPPopt that was associated with impaired PRx (>+0.15) was denoted as being below the lower limit of reactivity (LLR). Glasgow Outcome Scale (GOS) score was assessed at 6 months post-ictus. RESULTS: When ΔCPPopt was plotted against PRx and stratified by GOS groupings, data belonging to patients with a more unfavourable outcome had a U-shaped curve that shifted upwards. More time spent with a ΔCPPopt value below the LLR was positively associated with mortality (area under the receiver operating characteristic curve = 0.76 [0.68-0.84]). CONCLUSIONS: In a recent cohort of patients with severe TBI, the time spent with a CPP below the CPPopt-derived LLR is related to mortality. Despite aggressive CPP- and ICP-oriented therapies, TBI patients with a fatal outcome spend a significant amount of time with a CPP below their individualised CPPopt, indicating a possible therapeutic target.