A randomised controlled trial was undertaken to determine whether a respiratory health worker was effective in reducing the respiratory impairment, disability, and handicap experienced by patients with chronic airflow limitation attending a respiratory outpatient department. The 152 adults (aged 30-75 years) who participated had a prebronchodilator forced expiratory volume in one second (FEV1) below 60% predicted and no other disease. They were randomised to receive the care of a respiratory health worker or the normal services provided by the outpatient department. The respiratory health worker provided health education and symptom and treatment monitoring in liaison with primary care services. After one year there was little difference between the two groups in spirometric values (FEV1 and forced vital capacity before and after salbutamol 200 micrograms), disability (six minute walking distance and paced step test), and handicap (sickness impact profile, hospital anxiety and depression scale). Patients not looked after by the respiratory health worker were more likely to die (relative risk 2.9 (95% confidence limits 0.8, 10.2); when age and FEV1 were controlled for this risk increased to 5.5 (95% confidence limits 1.2, 24.5). Patients looked after by the respiratory health worker attended their general practitioner more frequently and were prescribed a greater range of drugs. This is the third study to have found limited measurable benefit in terms of morbidity from the intervention of a respiratory health worker. This may be due to the ability of such workers to keep frail patients alive.