A number of studies9–12 have found a dissociation between asthma severity based on symptoms and measures of lung function. For example, a study10 in children found that, when asthma severity was based on symptoms or medication use, there was no relationship between the NAEPP 1997 severity classification and FEV1. A reanalysis,13of data from two large, randomized controlled clinical trials (n =1,576) found large within-patient variability and no more than a moderate correlation between the changes in FEV1 and PEF, FEV1 and daily symptoms score, and daily symptom score and β2-agonist use. Considerable intrapatient variability in asthma severity classification over time was also demonstrated in a study14 of 85 steroid-naïve patients. At baseline, all patients met the criteria for moderate or severe persistent asthma. However, the mean percentage of the 12 treatment weeks when these patients would be reclassified as having intermittent, mild, moderate, and severe asthma was 9%, 14%, 71%, and 6%, respectively. This variability was dependent on changes in any one of the measures of severity: PEF, β2-agonist use, or symptoms.,14 The lack of correlation between the factors used in the NAEPP 1997 severity classification is of concern, as they can all be used to make the final severity assessment. In addition, the observed variability over time means that a severity classification may depend on which particular day that patient is assessed and their recent symptom history.