We measured carbon monoxide diffusing capacity of the lungs (DL,CO) by both the resting single-breath (SB) and steady-state (SS) exercise methods in 95 patients referred for pulmonary function testing. A 10-second breath-holding method was used for the SB test. DL,CO (SS) was measured during the last minute of a 3-minute exercise test on a 9-inch step. Results of the two methods showed good agreement, the SB-SS difference averaging -0.70 (SD, 3.39) ml/min per mm Hg. The difference between the two methods was not correlated with other measurements of pulmonary function except minute ventilation during the exercise performed in the DL,CO (SS) procedure. In a separate study of laboratory personnel, the day-to-day variabilities of the two tests were similar (SD, 1.4 ml/min per mm Hg). Alveolar volume obtained by helium dilution during the SB test was comparable to total lung capacity (TLC) estimated by multiple-breath nitrogen washout in patients without severe airway obstruction. In severe airway obstruction, the mean SB alveolar volume was 13.8% less than the TLC by nitrogen washout, a difference that may be useful as an indicator of inefficiency of gas mixing in the lungs. We conclude that the SB and SS exercise methods provide similar estimates of DL,CO in patients referred to a pulmonary function laboratory.