Although various methods have been used to correct the output of a respiratory mass spectrometer for the delay and rise time in its response, thereby reducing the error in measured gas concentrations, there are additional considerations in the case of a multiplex mass spectrometer. The measured signal from the detector is a series of discrete samples of the concentration of several different gases rather than a continuous monitor of a single gas concentration as generated by other types of mass spectrometer. If the time constant of the mass spectrometer is of the same order as the interval between samples of a gas in the multiplex mode, correction techniques based on continuous-time analysis would not be as valid as those based on discrete-time analysis. Such correction techniques were compared to the use of a simple time shift. For multibreath gas washout analysis with simulated worst-case 'square wave breathing' it was found that, because of the complex nature of the response of the mass spectrometer, a simple time shift provided a reduction in error nearly equal to that of an additional first order response correction, and that such further corrections may be unnecessary or even invalid under some circumstances.