Turbuhaler and Salbutamol-Diskus produce therapeutic doses at peak inspiratory flow (PIF) of >30 L/min. However, the optimum flow for Fluticasone-Diskus and Turbuhaler, in terms of total emitted dose and fine particle mass, is >60 L/min. The Turbuhaler achieved a higher output at this flow, as compared to Diskus. For pMDI 25 < PIF < 90 L/min, an actuation time of 0.0-0.2 sec is optimal. The aim of this study was to examine the incidence of optimum inhalation profiles, the effect of instruction, reproducibility, and the relationship between inhalation profiles and patient characteristics in stable asthmatics and mild/moderate/severe COPD patients. For each device, triplicate inhalation profiles were recorded during 6 sessions in a 10-week period. All patients achieved PIF > 30 L/min using Diskus. After instruction, all Diskus inhalations were performed with >60 L/min, except 7% of the inhalations of the severe COPD patients. At least 95% of the Turbuhaler inhalations was also performed with the minimum flow; however, 19% of the inhalations of the severe COPD patients were not optimally performed. The hand-lung coordination was inadequate in 40% of pMDI inhalation profiles, and 80% was performed with a too high flow. The reproducibility of PIF of both dry powder inhalers (DPIs) was very high (coefficient of variation = 4-10%). The reproducibility of the pMDI variables was lower (coefficient of variation = 9-18%). The major lung function variables predictive for PIF(diskus) and PIF(turbuhaler) were maximal inspiratory mouth pressure (MIP), PIF, and inspiratory capacity. No significant predictive lung function variables for PIF(pMDI) were found. Most patients performed reproducible optimum inhalation profiles through Diskus and Turbuhaler. However, in the severe COPD group, 7-19% of the patients were not able to generate the optimum flows through the DPIs. For these patients, a flow-independent aerosol delivery system might be more suitable. The majority of patients were using the pMDI incorrectly. Instruction had no effect. So, we concluded that the pMDI should not be used in these patient groups because of the coordination problems.