It is one of the unsolved mysteries in neurosurgery: why and when intracranial aneurysms rupture. Several causes have been adduced, like the size of the aneurysm, its location, tissue properties of the aneurysm wall and changes in transmural pressure gradients, just to name a few. Similarly, nobody knows definitely why an aneurysm stops bleeding once it has ruptured. In some instances it may be a temporary breakdown of cerebral circulation due to a massively increased intracranial pressure, while others believe that a local vasospasm and thrombus formation are functional in stopping the haemorrhage. Some more substantial ideas are available for the causes of re-haemorrhage, like thrombolysis and fibrinolysis in the vessel and around the rupture location, as well as relaxation of the local spasm. In this context, it is of utmost importance to know which factors could increase the risk of re-bleeding once the patient reached clinical service, and is suitable for treatment of the aneurysm. The paper of Fountas et al. has to be considered in this context. It reviews the literature regarding the relation of rebleeding and external ventricular drainage after aneurysmal subarachnoid haemorrhage (SAH). It is a well known fact that acute hydrocephalus after SAH can deteriorate the patient’s initial condition severely. In other words, a low—and thus favourable—Hunt Hess Grade can be masked by an acute hydrocephalus. Therefore, a timely drainage of a hydrocephalus can have a significantly positive influence on the patient’s clinical status. Thus, this therapeutic option is an essential part of the treatment of aneurysmal SAH and may be quite often necessary even before a definitive occlusion of the aneurysm is possible. In their study, the authors scanned the published patient series of the last 50 years. Unfortunately the results are ambiguous. A number of studies report an increased incidence of haemorrhages after external ventriculostomy in patients with unclipped aneurysms. The incidence ranges from 17% to 43% and is higher than in the control groups used in these studies. Some of the authors see a direct relation of the incidence of rebleeding to the level of intracranial pressure. Therefore, they recommend a special management of the ICP levels to reduce the risk of rebleeding. On the contrary to these results, Fountas et al. found several studies that could not see any influence of ventricular drainage on the incidence of re-haemorrhage at all. The rate of rebleeding ranges from 4.4% to 16% in these ‘negative’ studies. As Fountas et al. emphasise, the studies are very heterogeneous in their design. One major problem is the controls. Some studies use real controls; others compare their data with historical data. In addition, the comparison of the percentage of the re-haemorrhages reported in the ‘positive’ and ‘negative’ studies shows that there seems to be some overlap, as 16–17% does not show so much difference. In addition, the patient groups included in the studies are not homogenous, as patients with haemorrhage of unknown aetiology were included, as well. Another relevant parameter seems to be the duration of drainage, as well as the timing of aneurysm occlusion. As one could predict, the earlier the definitive occlusion of the aneurysm the less was the rebleeding rate, even in the patients with ventricular drainage. In the interpretation of the data some overlap of the natural rebleeding rate and the influence of ventriculostomy have also to be taken into account. Finally, it is well known that poor grade patients have higher incidences of hydrocephalus and rebleeding. These additional influences have been addressed clearly only in one paper. This commentary refers to the article http://dx.doi.org/10.1007/ s10143-005-0423-4