COPD is the most common chronic lung disease. It becomes more prevalent with increasing age but remains under-diagnosed in the elderly. A heated debate concerns the most suitable way to diagnose airway obstruction in this age group. Most COPD guidelines recommend to use a FEV1/FVC of 0.70 as threshold to define an obstructive ventilatory defect. While the use of a 0.70 ratio may be simple, it may result in under-diagnosis of airflow obstruction in young people and over-diagnosis in the elderly due to an age-related decline in pulmonary volumes, especially in FEV1. This might lead to unnecessary use of medications and increased risk of adverse effects. A proposed strategy for reducing the misclassification of airway obstruction include the use of the statistically derived lower limit of normal (LLN) for FEV1/FVC, calculated as the fifth percentile of the normal distribution in a healthy population. Some authors addressed the possibility to evaluate lung function through high-resolution CT. In contrast to spirometry, CT imaging may allow for regional assessment of the compartments involved (airways, parenchyma, vasculature), and may enable a phenotype-driven characterization of COPD. Bhatt et al. recently observed that a small proportion (7%) of subjects with CT-defined emphysema were identified by the 0.70 threshold for FEV1/FVC but not by the LLN. However, there is no evidence that CT-emphysema corresponds to a clinical entity that can benefit by inhaled therapy. Further studies are needed to assess the classificatory and prognostic value of the different proposed criteria to diagnose airway obstruction in the elderly.