Sir: At its meeting in Barcelona in September 1987, the Executive Committee of World Federation of Neurosurgical Societies received and unanimously approved a report from a committee that had been working for six years to devise a simple, reliable, clinically valid scale for grading patients with a subarachnoid haemorrhage. The committee's view was that the scale was needed for describing changes in an individual patient at different times, for estimating prognosis and for standardising assessment of management in different groups of patients; a requirement was that the scale would meld with currently used scaling systems. The committee took into account an analysis of data from the international cooperative aneurysm study which contained 3521 patients from 68 countries;' this showed that the two most important prognostic factors were the level ofconsciousness (important for the prediction of both death and disability) and the presence or absence of hemiparesis and/or aphasia (important only for disability in survivors). The analysis had shown that ifconsciousness was normal, headache and/or a stiff neck did not significantly affect outcome. The committee resolved that five grades only should be used for patients with subarachnoid haemorrhage; patients with an unruptured aneurysm should be identified separately, or classified as 'zero'. It believed that the Glasgow Coma Scale2 should be used to assess the level of consciousness, because of its world wide acceptance in assessment of coma from head injury. The only additional factor should be the presence or absence of major focal deficit to differentiate between grades two and three. The committee also resolved that in assessing outcome from subarachnoid haemorrhage, categories of the Glasgow Outcome Scale should be used3: Dead Vegetative Survival, Severely disabled; Moderately disabled; good recovery.