Three hours after spinal anaesthesia, a previously healthy 71-year-old man developed a lethal intracranial subarachnoid haemorrhage that was very likely due to rupture of a pre-existing aneurysm. Three weeks previously the patient had undergone an operation for which spinal anaesthesia had been administered uneventfully. The question is raised whether there was a causal relationship between the haemorrhage and the spinal anaesthesia; this is discussed in the light of similar cases previously published. Loss of cerebrospinal fluid (CSF) from the puncture site with a subsequent decrease in CSF pressure and an increase in transmural wall tension of the vessels might be predisposing factors for the rupture of a pre-existing cerebral aneurysm. Thus, we believe that intracranial subarachnoid haemorrhage should be listed among the rare complications of spinal anaesthesia. A dural leak following lumbar puncture might persist for months or even years without causing symptoms. In case of a planned second puncture, persisting leakage should be ruled out by taking a thorough history. Spinal and epidural anaesthesia are contraindicated in patients with persisting low pressure in the CSF system or known intracranial vascular malformations.