From a randomly selected population representative of the white population of Tucson, Ariz., satisfactory flow-volume data were obtained for 3,115 persons. Data from the 746 subjects who were totally free of symptoms or history of cardiorespiratory disease and who had never smoked were used in determining "normal" prediction equations for spirometric parameters and maximal expiratory flows. The maximal expiratory flow-volume curve showed considerable intersubject variability, but little change in shape of the mean maximal expiratory flow-volume curve was seen with advancing age when the effects of disease, insult, or injury were excluded.