Background: Chronic obstructive pulmonary disease (COPD) is a significant health problem worldwide. This randomised controlled trial aims at testing a new approach that involves a registered nurse working in partnership with patients, general practitioners (GPs) and other health professionals to provide care to patients according to the evidence-based clinical practice guidelines. The aim is to determine the impact of this partnership on the quality of care and patient outcomes. Methods: A cluster randomised control trial design was chosen for this study. Randomisation occurred at practice level. GPs practising in South Western Sydney, Australia and their COPD patients were recruited for the study. The intervention was implemented by nurses specifically recruited and trained for this study. Nurses, working in partnership with GPs, developed care plans for patients based on the Australian COPDX guidelines. The aim was to optimise patient management, improve function, prevent deterioration and enhance patient knowledge and skills. Control group patients received 'usual' care from their GPs. Data collection includes patient demographic profiles and their co-morbidities. Spirometry is being performed to assess patients' COPD status and CO analyser to validate their smoking status. Patients' quality of life and overall health status are being measured by St George's Respiratory Questionnaire and SF-12 respectively. Other patient measures being recorded include health service use, immunisation status, and knowledge of COPD. Qualitative methods will be used to explore participants' satisfaction with the intervention and their opinion about the value of the partnership. Analysis: Analysis will be by intention to treat. Intra-cluster (practice) correlation coefficients will be determined and published for all primary outcome variables to assist future research. The effect of the intervention on outcomes measured on a continuous scale will be estimated and tested using mixed model analysis of variance in which time and treatment group will be fixed effects and GP practice and subject nested within practice will be random effects. The effect of the intervention on the dichotomous variables (such as smoking status, patient knowledge) will be analysed using generalised estimating equations with a logistic link and a model structure that is analogous to that described above.