OBJECTIVES: The Assessment of Burden of COPD (ABC) tool evaluates and visualizes the health status of patients with chronic obstructive pulmonary disease (COPD). This tool may be used during consultations to monitor the burden of COPD and to adjust treatment. The ABC tool has items in 5 dimensions: symptoms, limitations, mental status, exacerbations, and fatigue. The aim of our study was to determine the burden of each of 15 elements in the ABC questionnaire. METHODS: A discrete choice experiment (DCE) was conducted using telephone-assisted personal interviews with 279 COPD patients. They were presented 13 sets of two patients, and decided which patient was in the worst health state. In order to make the choice task feasible despite the large number of 15 attributes, we generated an efficient partial profile design, that kept attributes in some dimensions constant and all at the same level (fold in), while varying the attributes in other dimensions (fold out). This reduced the burden on respondents by having them assess some attributes combined in categories, instead of as separate attributes. Multinomial logit was used to analyze the data. RESULTS: Patients were considered to be in worst health if they had high levels of fatigue, exacerbations, anxiety, breathlessness at rest and limitations in moderate physical activities. These factors had three to five times as much impact on the burden of disease as lower levels of these attributes and other limitations. Coefficients were very small and/or statistically insignificant for breathlessness during physical activity, limitations on strenuous activity, coughing, small numbers of exacerbations, and most mental attributes. CONCLUSIONS: It is possible to administer cognitively complicated DCE questionnaires using a fold-in-fold-out design. COPD-patients seem to accept being unable to perform strenuous activities. Gains in well-being can primarily be achieved by focusing on patients’ ability to lead a relatively normal everyday life.