BACKGROUND--The percentage of patients inhaling their medication effectively varies widely, according to methods of assessment and inhalers used. This study was carried out to assess differences among four types of inhalers using inhaler-specific checklists. METHODS--Inhalation technique was evaluated in adult patients with chronic obstructive pulmonary disease (COPD). Inhalers investigated were either metered dose inhalers (MDIs) or the dry powder inhalers Turbohaler (Turbuhaler), Diskhaler, and Rotahaler. Errors were recorded against inhaler-specific checklists. From these, scores were derived by dividing the number of items correctly completed by the total number of items on the checklist and the result was expressed as a percentage. For every inhaler "essential actions" were identified and scores on these key manoeuvres were calculated. The percentage of patients performing all these essential actions correctly was also calculated. Scores were also compared with adjustment for differences in relevant patient characteristics. RESULTS--Important differences among inhalers were found. Of 152 patients with COPD (mean (SD) age 55.1 (8.7) years), those with MDIs performed worst, especially when only essential items were considered. Patients with a Diskhaler did best, although after correction for patient characteristics the differences tended to diminish. Only 60% of patients were able to perform all essential inhaler actions satisfactorily. Of those using the Diskhaler, 96% did so correctly, while the corresponding figure for those using the MDI was only 24%. CONCLUSIONS--Many patients with COPD use their inhaler ineffectively. After adjusting for patient characteristics, differences among inhalers, although less pronounced, persist. Patients using a Diskhaler made fewest errors, while most patients using MDIs made crucial mistakes.