CENTRAL neuraxial blockades (CNBs) have gained widespread use, but may also be associated with serious complications. Spinal haematoma (SH) complicating CNB was initially believed to be an exceptional event. Indeed, based on published case reports, the first incidence of SH following CNB was reported as 1 : 1,000,000. However, several studies published in the last decades have found significantly higher incidences, particularly following non-obstetric epidural blockade (EB). Two large studies report the incidence of SH among non-obstetric perioperative EB as 1 : 10,000– 20,000. The probability for any single patient to develop SH following EB is very low, but the potentially disastrous consequences require attention for preventive measures, and timely diagnosis and management of SH is mandatory for full restitution of the patient. As all studies indicate that SH is extremely rare following spinal blockade (SB), it is remarkable that one study reported SH following SB in a patient subgroup of five females with fractured neck of femur, allowing an incidence of 1 : 22,000. The incidence of SH in the remaining perioperative SBs was 1 : 480,000. Female orthopaedic patients also appear to be at a particular high risk of developing SH following EB for prosthetic limb surgery, with an incidence of approximately 1 : 4000, reported from both sides of the Atlantic. Other risk factors seem to be vascular surgery, and renal and hepatic impairment. Disturbed coagulation, pharmacologically induced or otherwise caused, is also of major concern regarding the risk of developing SH. Removal of an epidural catheter is as critical a moment as its placement, and coagulation might have been impaired following surgery. Several studies have addressed the CNB’s potential of reducing morbidity and mortality, with conflicting results. Recent advances in patient care and optimal CNB techniques are not always applied. The individual risk–benefit analysis has to take this into account, integrating adequate CNB technique with all other aspects of optimised perioperative medical care. It is also possible that the patient with the largest potential benefit of perioperative EB is at a greatest risk of developing complications following the blockade. On the other hand, obstetric patients are at a significantly lower risk of developing SH following any kind of CNB. A metaanalysis reported the incidence of SH as 1 : 168,000 following obstetric EB, and this low incidence was recently confirmed by a large study that reported no case of obstetric SH among 186,900 EB (including 25,350 continuous spinal-EBs). Moreover, the majority of the rare obstetric SHs are caused by coagulation disorders related to specific obstetric conditions, such as preeclampsia, coagulopathy following large obstetric haemorrhage, or the syndrome of haemolysis, elevated liver enzymes and low platelets. In contrast, bloody taps, so frequently encountered when performing obstetric EB, apparently do not cause SH in obstetrics, but are often associated with cases of perioperative SH. Pregnancy itself induces a condition of increased coagulation, and NSAIDs are rarely used during pregnancy. Pharmacologically impaired coagulation is infrequent during pregnancy, and there are