Aims: Benefits of long acting beta 2 agonists are unclear for severe chronic obstructive pulmonary disease (COPD) patients with poor response to short acting bronchodilators. We aimed to evaluate 1) effects of eformoterol in such patients using a ‘n-of-1’ double crossover study design, and 2) aggregate data as a double-blind, double crossover randomized control trial. Methods: Subjects with forced expiratory volume in one second (FEV1) < 60% predicted, and poor response to short acting bronchodilators were studied six times over 18 weeks. During that time they were prescribed four weeks of either eformoterol or placebo, followed by the alternate, and then a second crossover. Four-weekly measures included six minute walk distance (6MWD), FEV1, previous two weeks of symptoms, and chronic respiratory questionnaire (CRQ) including treatment goal items. Results: Of 27 original subjects (21 male, mean age of 70 years, five smokers, mean prebronchodilator FEV1, 36% predicted), one subject had clinically significant concordant improvement in the CRQ dyspnoea domain and 6MWD (by 51 metres), but not for other outcomes. There were no concordant improvements in any other subjects. Aggregate double crossover data analysis demonstrated no improvement in any outcome measures. Conclusions: The ‘n-of-1’ study design and aggregate data analysis demonstrated lack of benefit from eformoterol in COPD patients with poor response to short acting bronchodilators.