In their study in this issue, Morris et al4 have gone one step further to differentiate, within a very large database, smoking and nonsmoking patients with normal ratios, patients with FEV3/FVC below LLN, and patients with both FEV3/FVC and FEV1/FVC below LLN, comparing total lung capacity, residual volume/total lung capacity, inspiratory capacity, and diffusing capacity, according to LLN. The main finding of the study was that a group of patients had an FEV3/FVC ratio below LLN and otherwise normal measures. These measures were, however, significantly different from those in the group with both normal FEV1/FVC and FEV3/FVC and from the group defined as obstructive according to FEV1/FVC below LLN. They were more likely to be smokers. Because the variability in FEV3/FVC is very low in healthy subjects (so that deviations from the predicted values are very small, smaller in fact than for FEV1/FVC),7 there is some rationale for the potential usefulness of the FEV3/FVC ratio to detect mild disease. This group represented only 10% of all patients with obstruction. The finding may simply reflect a statistical bias because LLN at the 95% confidence level for each of the nine measured parameters led a number of otherwise normal subjects to become abnormal for one of these parameters; however, 10% is above the expected bias level. The group was significantly older than the normal group but not older than the obstructive group. This finding potentially supports the hypothesis that the obstructive disease was progressing more slowly or had appeared later. No detailed data on the duration or intensity of cigarette smoking is provided. This small group might include otherwise normal patients with a borderline accelerated aging process translating into an isolated lower FEV3/FVC ratio, as one-third of the patients with isolated significant FEV3/FVC were nonsmokers. Indeed, correlation with age is higher for FEV3/FVC than for FEV1/FVC and current smokers have even further significant decrease,5 but the analysis of the data according to LLN was corrected for age in the study. Finally, if we accept the convincing evidence that such a group has been characterized, then we need to understand why this intermediate group—between health and disease according to current standards—is such a small percentage of the population. FEV3/FVC might in fact be only a little less insensitive to obstruction than FEV1/FVC or not detect exactly the same kind of disease. Or could it be that the transition phase between early (mild) unrecognized airway obstruction and diagnosed obstruction takes less time than we believe?