The diffusing capacity (D L,CO) has been measured repeatedly by an improved “brief steady-state” technique in a total of 13 experimental and 4 control subjects. Both resting and exercise (D L,CO) were initially somewhat higher in the more athletic subjects. However, physical training sufficient to increase the predicted maximum oxygen intake by an average of 12% over a 6–15 week period was without significant influence upon (i) the restingD L,CO (whether standardized for respiratory rate, tidal volume, or respiratory minute volume) and (ii) exerciseD L,CO at the mean oxygen consumption of the individual. When account was taken of the increase in maximum oxygen intake, there was some increase in the predictedD L,CO max (< 5%); however, this was too small to be of great practical significance. Existing evidence for an increase of restingD L,CO with training is shown to be insubstantial. The underlying hypothesis thatD L,CO limits oxygen transfer is also weak; if a small increase ofD L,CO does occur during training, this is by virtue of the close association betweenD L,CO and pulmonary blood flow.