measurements, as well as acceptable measurements in more than 1,000 children aged 5-12 years. A comparison of children living in a range of locations in India with those of Indian heritage who were living in London in the United Kingdom revealed significantly higher FEV 1 and FVC (per z-scores relative to reference values derived from Indian populations) in the London residents. However, when the India-resident children were subgrouped into urban, semiurban, and rural areas of residence, it was discovered that the urban-residing children from India had spirometric measures that were indistinguishable from those measured in the London contingent. In contrast, the rural-residing Indian children had FEV 1 and FVC measures that were 0.9 z-scores lower, and the semiurban Indian group had measures 0.5 z-scores lower, than the India urban and London groups. Assuming that the genetically determined physical characteristics relating predicted lung volumes to height, age, and sex are relatively constant for the general population of India, these results suggest strongly that differences in living conditions among these subpopulations have a significant effect on lung development or lung function. This is useful information, as it identifies a population that could hold clues to mechanisms by which environmental conditions may affect lung development in utero or in early life and that may benefit from interventions to improve lung health.The authors also report that the z-FEV 1 /FVC was not significantly different among the subpopulations, suggesting a restrictive (reduced lung size), rather than obstructive, pattern in the nonurban residents. This needs to be confirmed with a measure of total lung capacity (TLC), which would not be possible for population screening but could be attempted in a representative sample of participants. In persons with normal respiratory systems, the primary determinant of FEV 1 is FVC, the primary determinant of FVC is TLC, and the primary determinant of TLC is thoracic volume, which relates to height and chest dimensions. Thus, the surrogate measures (height, age, sex, ethnic heritage) that are used to compute reference ranges for spirometry mostly function as an estimate of TLC. If the actual TLC is significantly different from the expected TLC, this will affect the z-FEV 1 and z-FVC but will not affect the z-FEV 1 /FVC unless the TLC is altered by a process that also affects airway integrity or lung elastic recoil. The alternative scenario that may cause a proportional decrease of FEV 1 and FVC is premature airway closure during forced expiration (2), although this is unlikely in the absence of chronic respiratory symptoms.The authors employed a multivariate model that included family affluence score, exposure to household tobacco smoke, exposure to indoor biomass smoke, z-body weight, and z-body mass index, but these variables were nearly qualitatively associated with place of residence, so none emerged as strong correlates with the spirometric measures. Specific data regarding maternal to...