The Glasgow Coma Scale (GCS) is the most universally accepted system for grading level of consciousness. Predicting outcome is particularly difficult in poor grade aneurysmal subarachnoid haemorrhage (aSAH) patients. We hypothesised that the GCS and individual examination components would correlate with long-term outcome and have varying prognostic value depending on assessment time points. GCS scores of 160 aSAH patients presenting in stupor or coma were prospectively recorded on admission and each subsequent day until hospital day 14. Early treatment was planned for each patient unless the patient's family refused aggressive intervention or the patient died before surgery. Outcomes were assessed by the modified Rankin scale (mRS) at 14 days, 3 months, and one year. All patients who did not receive surgical treatment died within one year. Of the 104 patients who received surgical treatment, 13.5% of them had a favourable outcome at 14 days, 38.5% at 3 months, and 51% at one year (p<0.0001). Admission GCS scores significantly correlated with outcome (Spearman rank test, rs=0.472, p<0.0001). On admission, motor examination correlated best with one-year outcome (rs=0.533, p<0.0001). Each point increase in motor examination predicted a 1.8-fold increased odds of favourable long-term outcome (95% confidence interval [CI], 1.4-2.3). At discharge, eye examination (rs=0.760, p<0.0001) correlated best with one-year outcome, and a one point increase in eye examination predicted a 3.1-fold increased odds of favourable outcome (95% CI, 1.8-5.4). During hospitalisation, the best eye exam (rs=0.738, p<0.0001) and worst motor exam (rs=0.612, p<0.0001) were the most highly correlated with the one-year outcome. Long-term follow-up is necessary when evaluating recovery after aSAH, as outcomes improve significantly during the first year. The GCS and its individual components correlate well with long-term outcome. Admission motor examination and spontaneous eye opening during hospitalisation are most predictive of favourable recovery.