In 14 normal subjects and in 13 patients with obstructive pulmonary diseases, we studied the variability within an individual of values for the maximal expiratory flow rate (Vmax) recorded simultaneously vs expired pulmonary volume (at the mouth) and vs thoracic volume (measured with a body plethysmograph). We found that the variance of Vmax within an individual at 25, 50, and 75 percent of the expired vital capacity did not differ statistically whether pulmonary volume was the expired or the thoracic gas volume. In ten healthy subjects on two occasions (at an interval of 12 days, on the average), we measured the peak expiratory flow rate and Vmax at different levels of inflation, with respect to either expired or thoracic volume. There was no statistical differences in Vmax between the first and the last day. A larger variability of Vmax measured vs expired volume implies a change in the expiratory effort from one forced expiration to another and a different degree of compression of intrathoracic air. Since this was not the case, we conclude that muscular effort during repeated forced expirations is similar. The good reproducibility of effort explains in great measure the good reproducibility of Vmax.