Clinical coding is the translation of documented clinical activities during an admission to a codified language. Healthcare Resource Groupings (HRGs) are derived from coding data and are used to calculate payment to hospitals in England, Wales and Scotland and to conduct national audit and benchmarking exercises. Coding is an error-prone process and an understanding of its accuracy within neurosurgery is critical for financial, organizational and clinical governance purposes. We undertook a multidisciplinary audit of neurosurgical clinical coding accuracy. Neurosurgeons trained in coding assessed the accuracy of 386 patient episodes. Where clinicians felt a coding error was present, the case was discussed with an experienced clinical coder. Concordance between the initial coder-only clinical coding and the final clinician-coder multidisciplinary coding was assessed. At least one coding error occurred in 71/386 patients (18.4%). There were 36 diagnosis and 93 procedure errors and in 40 cases, the initial HRG changed (10.4%). Financially, this translated to pound111 revenue-loss per patient episode and projected to pound171,452 of annual loss to the department. 85% of all coding errors were due to accumulation of coding changes that occurred only once in the whole data set. Neurosurgical clinical coding is error-prone. This is financially disadvantageous and with the coding data being the source of comparisons within and between departments, coding inaccuracies paint a distorted picture of departmental activity and subspecialism in audit and benchmarking. Clinical engagement improves accuracy and is encouraged within a clinical governance framework.