After traumatic brain injury, decompressive craniectomy (DC) is a second-tier, late therapy for refractory intracranial hypertension. We hypothesize that early DC, based on CT evidence of intracranial hypertension, improves intracranial pressure (ICP) and cerebral perfusion pressure (CPP). From September 2008 to January 2015, 286 traumatic brain injury patients requiring invasive ICP monitoring at a single Level I trauma center were reviewed. DC and non-DC patients were propensity score matched 1:1, based on demographics, hemodynamics, injury severity score (ISS), Glasgow Coma Scale (GCS), transfusion requirements, and need for vasopressor therapy. Data are presented as M ± SD or median (IQR) and compared at P ≤ 0.05. The study population was 42 ± 17 years, 84 per cent male, ISS = 29 ± 11, GCS = 6(5), length of stay (LOS) = 32(40) days, and 28 per cent mortality. There were 116/286 (41%) DC, of which 105/116 (91%) were performed at the time of ICP placement. For 50 DC propensity matched to 50 non-DC patients, the midline shift was 7(11) versus 0(5) mm (P < 0.001), abnormal ICP (hours > 20 mm Hg) was 1(10) versus 8(16) (P = 0.017), abnormal CPP (hours < 60 mm Hg) was 0(6) versus 4(9) (P = 0.008), daily minimum CPP (mm Hg) was 67(13) versus 62(17) (P = 0.010), and daily maximum ICP (mm Hg) was 18(9) versus 22(11) (P < 0.001). However, LOS [33(37) versus 25(34) days], mortality (24 versus 30%), and Glasgow Outcome Score Extended [3.0(3.0) versus 3.0(4.0)] did not improve significantly. Early DC for CT evidence of intracranial hypertension decreased abnormal ICP and CPP time and improved ICP and CPP thresholds, but had no obvious effect on the outcome.