Background: Obtaining informed consent for a patient’s participation in a randomized trial of treatment for use in a medical emergency may be achieved in a variety of ways. We sought to assess the process of consent and to evaluate the influence of the patient’s neurological deficit on the method used to obtain consent in the first 300 patients recruited into the Third International Stroke Trial (IST-3). Methods: IST-3 is the first large-scale randomized controlled trial of intravenous thrombolysis in acute ischaemic stroke. The clinician could use one of four procedures to recruit the patient: written consent, witnessed consent, assent, or a waiver of consent. The patient’s neurological deficits were recorded at baseline. We analysed the relationship between the neurological deficits at baseline and the consent procedure. Results: The method of consent used was written consent in 71 subjects (24%), witnessed verbal consent in 30 subjects (10%), assent by a relative in 197 subjects (66%), and waiver of consent in 2 subjects (1%). Patients with severe neurological deficits (as measured either by their stroke syndrome or their lower predicted probability of being alive and independent at 6 months) were more likely to be recruited by assent. Patients able to give written consent had less severe strokes. Conclusions: Patients with non-lacunar hemispheric stroke syndromes or with a more severe neurological deficit were less likely to give written consent. Excluding such patients from acute stroke treatment trials would eliminate many otherwise eligible subjects, who have a poor predicted outcome without treatment and yet might benefit from acute treatments such as thrombolysis. Flexible consent procedures developed for IST-3 have made it feasible to recruit the target population. Copyright (c) 2006 S. Karger AG, Basel.