This editorial reviews the evidence behind the following hypotheses. (1) People with disabilities often do not receive the input from health care services they would benefit from and should expect. (2) Hospitals managing problems arising from disease also have increasing problems managing the disabilities of their patients, with recovery and discharge being prolonged. (3) Rehabilitation services are poorly organised and provided, with much provision now being located in separate specialist centres or delivered by separate expert services. (4) Yet there is good evidence of the effectiveness of rehabilitation. (5) This unsatisfactory situation arises from basing service organisation and delivery upon the biomedical model of illness, the culturally dominant model in most countries. The risks of continuing this separation are outlined.