The most commonly used technique for the measurement of pulmonary diffusing capacity (DL) is the single-breath hold technique requiring a 10-s breath-hold after the maximal inspiration of carbon monoxide (0.3% CO) and helium (10% He). To measure pulmonary diffusing capacity in our experiments, we had the added advantage of the use of the Gould Pulmonary Function Laboratory that automates the collection and recording of data and the calibration of equipment for each test. However, this technique, DL(CO), is difficult to use during exercise of moderate or elevated intensity because of the lengthy breath-hold. Thus, the purpose of the present study was to compare DL(CO) with 3-s and 5-s breath-holds to a 10-s breath-hold at rest and during moderate and intense exercise in 14 subjects. As expected, an increase in the DL(CO) was observed during moderate and intense exercise when compared to resting values (45.7 +/- 10.0 and 53.0 +/- 7.6 vs 32.1 +/- 7.7 ml CO min-1 mmHg-1). No difference was observed between values for DL(CO) measured at varying breath-hold times at rest (3 s: 32.9 +/- 7.4; 5 s: 32.0 +/- 7.5; 10 s: 31.4 +/- 8.2 ml CO min-1 mmHg-1) or during moderate exercise (3 s: 45.9 +/- 10.1; 5 s: 45.9 +/- 10.6; 10 s: 45.2 +/- 10.4 ml CO min-1 mmHg-1) or intense exercise (3 s: 52.1 +/- 8.3; 5 s: 54.3 +/- 9.3; 10 s: 52.6 +/- 5.2 ml CO min-1 mmHg-1). Reliability coefficients indicated that the use of a 3-s breath-hold was appropriate.(ABSTRACT TRUNCATED AT 250 WORDS)