AbstractClinical guidelines indicate that a chronic obstructive pulmonary disease (COPD) diagnosis is made from a single spirometry test. However, long-term stability of diagnosis based on forced expiratory volume in 1 s over forced vital capacity (FEV₁/FVC) ratio has not been reported. In primary care subjects at risk for COPD, we investigated shifts in diagnostic category (obstructed/non-obstructed). The data were from symptomatic 40+ years (ex-)smokers referred for diagnostic spirometry, with three spirometry tests, each 12±2 months apart. The obstruction was based on post-bronchodilator FEV₁/FVC < lower limit of normal (LLN) and <0.70 (fixed ratio). A total of 2,352 subjects (54% male, post-bronchodilator FEV₁ 76.5% predicted) were studied. By LLN definition, 32.2% were obstructed at baseline, but 32.2% of them were no longer obstructed at years 1 and/or 2. By fixed ratio, these figures were 46.6 and 23.8%, respectively. Overall, 14.3% of subjects changed diagnostic category by 1 year and 15.4% by 2 years when applying the LLN cut-off, and 15.1 and 14.6% by fixed ratio. Change from obstructed to non-obstructed was more likely for patients with higher body mass index (BMI) and baseline short-acting bronchodilator (SABA) users, and less likely for older subjects, those with lower FEV₁% predicted, baseline inhaled steroid users, and current smokers or SABA users at year 1. Change from non-obstructed to obstructed was more likely for males, older subjects, current smokers and patients with lower baseline FEV₁% predicted, and less likely for those with higher baseline BMI. Up to one-third of symptomatic (ex-)smokers with baseline obstruction on diagnostic spirometry had shifted to non-obstructed when routinely re-tested after 1 or 2 years. Given the implications for patients and health systems of a diagnosis of COPD, it should not be based on a single spirometry test.