As part of the ongoing Harvard Six Cities study of the respiratory effects of air pollution, repeated measurements of pulmonary function (FVC, FEV75, MMEF, and Vmax75) were taken at approximately weekly intervals in a population of 154 school children living in Kingston and Harriman, Tennessee. A series of as many as six measurements were obtained for each child over approximately a 2-month period beginning in February 1981. Concurrent measurements of ambient ozone, and fine particle and fine sulfate concentrations were obtained at a site near the study community. The maximal hourly ozone concentration observed during the study was 78 ppb. Child-specific linear regressions were fit that related short-term pulmonary function changes to air pollution or temperature. We found that decrements in FVC, FEV75, MMEF, and Vmax75 were associated with ozone, but not with particulate levels. Decrements in FVC, MMEF, and Vmax75 were also associated with temperature. Although slopes of pulmonary function on ozone varied across children, in general there was no evidence that this variation represented heterogeneity of response as opposed to random estimation errors. In addition, no evidence was found that individual response to ozone was related to sex, presence of asthma, respiratory illness before 2 yr of age, or the MMEF/FVC ratio, a rough surrogate for airway size. We conclude that ambient exposures to ozone at levels well below the National Ambient Air Quality Standard of 120 ppb are associated with transient decreases in lung function, the long-term significance of which is uncertain.