A comparison of measured and predicted signal‐to‐noise ratio is made in an optical heterodyne spectrometer utilizing a tunable diode laser which exhibited excess noise. It is shown that good agreement between predicted and measured signal‐to‐noise ratios results if excess noise effects due to tunable diode lasers are included in the predictions. The methods used to quantify excess noise and to incorporate their effects into predictions of signal‐to‐noise ratio are discussed.