Objective: To determine factors and outcomes associated with ultra-early surgery for poor-grade aneurysmal subarachnoid haemorrhage (aSAH). Design: A multicentre retrospective analysis, observational study. Setting: High-volume teaching hospitals (more than 150 aSAH cases per year). Participants: 118 patients with World Federation of Neurosurgical Societies (WFNS) grades IV and V underwent surgical treatment. Ultra-early surgery was defined as surgery performed within 24 h of aSAH, and delayed surgery as surgery performed after 24 h. Outcome was assessed by modified Rankin Scale (mRS). The mean time of follow-up was 12.5 +/- 3.4 months (range 6-28 months). Results: 47 (40%) patients underwent ultra-early surgery, and 71 (60%) patients underwent delayed surgery. Patients with WFNS grade V (p=0.011) and brain herniation (p=0.004) more often underwent ultra-early surgery. Postoperative complications were similar in ultra-early and delayed surgery groups. Adjusted multivariate analysis showed the outcomes were similar between the two groups. Multivariate analysis of predictors of poor outcome, ultraearly surgery was not an independent predictor of poor outcome, while advanced age, postresuscitation WFNS V grade, intraventricular haemorrhage, brain herniation and non-middle cerebral artery (MCA) aneurysms were associated with poor outcome. Conclusions: Although patients with WFNS grade V and brain herniation more often undergo ultra-early surgery, postoperative complications and outcomes in selected patients were similar in the two groups. Patients of younger age, WFNS grade IV, absence of intraventricular haemorrhage, absence of brain herniation and MCA aneurysms are more likely to have a good outcome. Ultra-early surgery could improve outcomes in carefully selected patients with poor-grade aSAH.