Available neurosurgical intensive care beds are rare and are often regarded as a precondition for accepting remote neurosurgical emergencies for operative treatment. From a medical and ethical point of view, the practice of rejecting acute patients as a result of a shortage of intensive care unit beds is questionable, because the primary need is fast treatment, and postoperative intensive care is secondary. We performed a retrospective analysis of outcome of 36 patients, who returned straight from our operating room to the intensive care unit of the referring hospital because no domestic neurosurgical intensive care unit beds were available (group 1). For the two largest cohorts, acute subdural haematoma (n=12) and primary intracerebral haemorrhage (n=15), outcome was compared by matched pair analysis with control patients kept in-house postoperatively (group 2). The mortality rate was higher in group 1 compared with group 2 (acute subdural haematoma 67 versus 58%; intracerebral haemorrhage 47 versus 40%). Despite this tendency, we regard 'operate-and-return' treatment as a sound and practicable alternative to offer to hospitals trying to refer patients with neurosurgical emergencies. It is to be preferred to denying treatment, because this results in a delay or cancellation of treatment, with the associated negative effects on outcome. However, postoperative neurosurgical intensive care unit treatment has to be considered the optimum goal, and there needs to be a continuous demand to authorities to increase the number of neurosurgical intensive care unit beds.