Following subarachnoid haemorrhage the most significant complication is sustained cerebral vascular contraction (vasospasm), which may result in terminal brain damage from cerebral infarction. Despite this, the biochemical cause of vasospasm remains poorly understood. In this study, the global high-concentration metabolite composition of CSF has been correlated with patient outcome after subarachnoid haemorrhage using multivariate statistics and 1H NMR spectroscopy. In total, 16 patients with aneurysmal subarachnoid haemorrhage (aSAH) were compared with 16 control patients who required a procedure where CSF was obtained but did not have aSAH. Multivariate statistics readily distinguished the aSAH group from the heterogeneous control group, even when only those controls with blood contamination in the CSF were used. Using principal components analysis and orthogonal signal correction, vasospasm was correlated to the concentrations of lactate, glucose and glutamine. These pattern recognition models of the NMR data also predicted Glasgow Coma Score (54% within +/- 1 of the actual score on a scale of 1-15 for the whole patient group), Hunt and Hess SAH severity score (88% within +/- 1 of the actual score on a scale of 1-5 for the aSAH group) and cognitive outcome scores (78% within +/- 3 of the actual score on a 100% scale for the whole patient group). Thus, the approach allowed the prediction of outcome as well as confirming the presence of aSAH.