We tested the hypothesis that quinolinic acid, a tryptophan-derived N-methyl-D-aspartate agonist produced by macrophages and microglia, would be increased in CSF after severe traumatic brain injury (TBI) in humans, and that this increase would be associated with outcome. We also sought to determine whether therapeutic hypothermia reduced CSF quinolinic acid after injury. Samples of CSF (n = 230) were collected from ventricular catheters in 39 patients (16 to 73 years old) during the first week after TBI, (Glasgow Coma Scale [GCS] < 8). As part of an ongoing study, patients were randomized within 6 hours after injury to either hypothermia (32 degrees C) or normothermia (37 degrees C) treatments for 24 hours. Otherwise, patients received standard neurointensive care. Quinolinic acid was measured by mass spectrometry. Univariate and multivariate analyses were used to compare CSF quinolinic acid concentrations with age, gender, GCS, time after injury, mortality, and treatment (hypothermia versus normothermia). Quinolinic acid concentration in CSF increased maximally to 463 +/- 128 nmol/L (mean +/- SEM) at 72 to 83 hours after TBI. Normal values for quinolinic acid concentration in CSF are less than 50 nmol/L. Quinolinic acid concentration was increased 5- to 50-fold in many patients. There was a powerful association between time after TBI and increased quinolinic acid (P < 0.00001), and quinolinic acid was higher in patients who died than in survivors (P = 0.003). Age, gender, GCS, and treatment (32 degrees C versus 37 degrees C) did not correlate with CSF quinolinic acid. These data reveal a large increase in quinolinic acid concentration in CSF after TBI in humans and raise the possibility that this macrophage-derived excitotoxin may contribute to secondary damage.