Diversion of suspected traumatic brain injury (TBI) patients to trauma centres may improve outcomes by expediting access to specialist neurosurgical care. This study aimed to determine the accuracy of the Head Injury Straight to Neurosurgery (HITSNS) triage rule for identifying patients with significant TBI. A diagnostic cohort study was performed using data from the HITSNS trial, the Trauma Audit and Research Network registry and the North East Ambulance service database. Sensitivity and specificity of the HITSNS triage rule were calculated against a reference standard of significant TBI, defined by a cranial Abbreviated Injury Scale score of at least 3 or by the performance of a neurosurgical procedure. A total of 3628 patients were included in the complete case analyses. The HITSNS triage tool demonstrated a sensitivity of 28.3% (95% confidence interval 21.8'35.4) and a specificity of 94.4% (95% confidence interval 93.6'95.2). The low sensitivity of the HITSNS triage rule suggests that a considerable proportion of patients with significant TBI may not be triaged directly to trauma centres, and further research is needed to improve the accuracy of bypass protocols.