OBJECTIVES: to test the efficacy of two educational interventions to improve the inhalation techniques in patients with COPD and the influence of patient´ preference. METHODS: Design: Multicenter patients´ preference trial or comprehensive cohort design ISRCTN15106246. Patients: 465 COPD patients (to detect a difference between groups of 25%, 80% statistical power, 95% confidence level, 40% expected losses), with inhaled treatment, written consent. Non-probabilistic consecutive sampling. Allocation: Patients without strong preferences for a treatment are randomised: RCT group (block randomization), and those with strong preferences are given their choice: PPS group. Variables: Primary outcomes: Performance of correct inhalation technique. Secondary outcomes: Pick flow, dyspnea (Baseline dyspnea index), Functional status (forced spirometry). Interventions: Intervention-A: Written information. A leaflet with the correct inhalation technique for the main inhaler devices used in our area. Intervention-B: Intervention-A + individual training (by instructors). Follow-up: 3 month, visits: baseline, 1 month, 3 month. Statiscal analysis: Mean, frequency, 95% confidence interval at baseline. Number Needed to Treat for a benefit (NNT) was calculated. Intention to treat analysis. RESULTS: Predominance of males (91.4%), mean age 69.8 years (CI95%, 69.00-70.59); FEV1(mean)=55.91% (IC95%, 53.62-58.2), mixed respiratory pattern (65.9%). Severity stage: 15.7% mild, 44.1% Moderate, 40.3% Severe. Pharmacological treatment: inhaled-beta2-adrenergic (88.8%); inhaled-corticosteroids (76.7%); inhaled-anticholinergic (70.7%); mucolitycs (19.4%); xanthine (7.3%); oral-corticosteroids (1.3%). BDI: grade 2. Primary outcome:  better inhalation technique (p=0.002) in the PPS group, NNT=7.4 (IC95%, 4.52-20).Among the RCT cohorts: there was no difference between control and intervention A and there were statistically significative differences between intervention B versus control (p<0.0001), NNT=2.44 (IC95%, 1.87-3.5) and  versus intervention A, NNT=2.85 (IC95%, 2.08-4.56). In the PPS cohorts: there was a difference (p<0.0001) between intervention B versus intervention A, NNT=2.33 (IC95%, 1.5-3.2). CONCLUSIONS: The performance of a correct inhalation Technique improves with monitor training. The patients´ preferences enhance the efficacy of intervention.