Brain–body interfaces (BBIs) have been shown through a number of studies to be useful assistive technology devices for recreation and communication. However, severely motor impaired persons with no other means of interacting with their environment have had difficulties using the standard communication software for the Cyberlink, the commercially available BBI device which we used in our studies. We have therefore developed a simple Yes/No program, drawing on a range of design and evaluation approaches from Human–Computer Interaction research. This paper presents the first extensive evaluation of this program. Its evaluation combines formal assessments with observations from users, carers and technology and mental health professionals who are involved in the research. Our conclusions are that the performance of severely impaired individuals cannot be readily separated from that of novice unimpaired users, that worthwhile results can be achieved with the Yes/No program, but at the moment the cost of using a Cyberlink is too high for most assistive technology contexts. However, for severely impaired individuals, Cyberlink use may be the only form of recreation and communication available to them, and thus the current limitations of the technology are acceptable for this user population.

We summarise the results of the first studies to investigate the Cyberlink brain body interface as an assistive technology. Three phases of studies and a contextual inquiry were performed with a range of users. A focus group was formed from brain-injured users with locked-in syndrome who have no other method of communication or control of a computer than the Cyberlink. Versions of a Yes/No program were then created to allow communication and have achieved some success with the focus group. The purpose of this paper is to discuss how this program has been improved and what steps need to be taken to create communication programs for persons with severe motor impairment. As a result of our experiences, we have been able to develop a set of design guidelines for brain-body interface operated Yes/No programs. These are presented and justified on the basis of our experiences. We also raise some general issues for assistive technologies of this nature.