The correct use of inhalation devices is essential to ensure the effectiveness of the treatment (Giraud and Roche, Eur Respir J 2002). We have shown in a large study published in 2003 that many patients make handling errors in real life (Molimard et al., J Aerosol Med 2003, Girodet et al., Therapie 2003). Since this study, new devices have been marketed in chronic obstructive pulmonary disease (COPD), including Breezehaler® and Respimat®. Interestingly, little is known regarding their real-life use. The aim of this study is to update data on device handling, including new devices, in a real-life setting in France. The primary objective is to evaluate real-life COPD patient handling of their usual inhaler (Breezhaler®, Diskus®, Handihaler®, Respimat®, Turbuhaler® or pressurized Metered Dose Inhaler) by general practitioners or pulmonologists.For this cross-sectional observational study, 1 000 GPs and 200 pulmonologists randomly selected from a geographically representative database, and who agree to participate in this study, will assess patient handling of their current inhaler. Five thousands patients will be requested to take a puff of their usual inhaler using their usual technique, which will be observed and rated by the physician. Patient characteristics, COPD history, and device use history, will be collected using a questionnaire. For each device (if a patient has several devices, each device will be studied) descriptive statistical analysis will be performed: the handling error rate will be estimated overall (handling errors will be classified as those independent or dependent of the inhalation system, and among the latter major/critical errors will be identified), as well as stratified according to patient characteristics, COPD history, device use history, and patient-reported treatment effect. Sensitivity analyses will be performed excluding patients having been trained within 3 months.