Objectives: The objective of this randomised, cross-over study was to compare a new single-dose dry powder inhaler (Elpenhaler (EH)), with a widely used, multi-dose dry powder inhaler (Diskus (DK)) on critical errors, patient preference, and satisfaction with the inhalers. Methods: First, patients read the instructions of one device, followed by a first inhalation attempt. Inhalation errors were assessed and if mistakes were made, correct inhaler use was demonstrated. Then patients had to demonstrate again and mistakes were registered. This was repeated up to four times. After completing the first device, the same procedure was started with the second inhaler. Primary outcome was the percentage of patients making at least one critical error after reading the insert. Secondary outcomes were inhaler preference and satisfaction with the inhalers. Results: After reading the insert, 19 of 113 patients (17%) made at least one critical error with DK and 40 (35%) with EH (p = 0.001); 73% preferred the DK and 27% the EH (p < 0.001). The mean overall satisfaction score (1 = very satisfied; 5 = very dissatisfied) for DK was 1.59 and for EH 2.48 (p < 0.001). Conclusion: With DK fewer errors were made, more patients preferred DK over EH and patients were more satisfied with DK. This may enable DK to improve treatment outcomes more than EH.