A formula for the signal‐to‐noise ratio of a maser spectrometer is derived by considering it as a special case of a cavity spectrometer. This formula is consistent with evaluations made by previous authors but is more convenient for comparison with an absorption cavity spectrometer. In applications where high resolution is not a requirement, the pressure and power level in an absorption spectrometer may be made very large so that its sensitivity can be superior to that of the maser. However, if it is operated to obtain the highest possible resolution, its sensitivity may or may not be superior to that of the maser, depending upon the frequency and upon other conditions. In the situation of greatest interest, in which the linear dimensions of the cavity are scaled in proportion to the wavelength and in which it is sufficiently large to make the effect of collisions between the molecules and the walls negligible, it is shown that the sensitivity of the maser relative to the absorption spectrometer varies inversely with the frequency. The theory is illustrated by calculations pertaining to spectrometers designed for the ND3inversion spectrum at 1500 Mc. From these calculations it can be inferred that, with the best available techniques, the two instruments would have about equal sensitivity in the region of 1500 to 2000 Mc, while at lower frequencies the maser would have higher sensitivity and the absorption spectrometer would have higher sensitivity at higher frequencies.