The measurement of single-breath diffusing capacity of the lung for carbon monoxide (DLCOsb) has traditionally used slow-responding analyzers that require large sample volumes for gas analysis. In addition, these measurement systems must use only assumed values of washout volume (Vwash) since gases are not measured continuously. More recently, a rapid analyzer was developed to continuously measure CO and methane (CH4, an inert tracer gas), which offers advantages, including accurate determination of Vwash and utilization of small sample volume (Vsamp) for gas analysis. We used this analyzer on 135 consecutive measurements of DLCOsb obtained from 70 patients and compared two analysis methods. (1) American Thoracic Society (ATS) guideline method: a standard Vwash (ATS-Vwash) of 0.75 L and a Vsamp of 0.75 L (for vital capacities [VC] less than 2.0 L, ATS-Vwash of 0.5 L and a Vsamp of 0.5 L); (2) VIS method: a Vwash visually just adequate to clear phases I and II of the CH4 washout curve (VIS-Vwash) and a Vsamp of 0.75 L (0.5 L for VC less than 2.0 L). We found that in four patients with VC less than 1.2 L, DLCOsb was attainable despite the small Vsamp (less than 500 ml). Of the other 66 patients whose tests were analyzed by both methods, 33 (50.0%) had Vis-Vwash less than or equal to ATS-Vwash such that phases I and II were actually cleared using ATS criteria. The other 33 patients (50.0%) had VIS-Vwash greater than ATS-Vwash such that phases I and II could not be cleared using ATS criteria.(ABSTRACT TRUNCATED AT 250 WORDS)