Rice first showed the theoretical relationship between the expected number of zero crossings over a unit time interval and the spectrum of stationary Gaussian random noise. If an additional signal causes the spectral content to change, then the number of zero crossings may change. It is the purpose of this paper to familiarize the reader with the recognition of the change in the zero‐crossing rate as a possible detection method. The principle of the effect is illustrated by two examples, one using a simple artificial noise mixture and the other using a noise recording made at sea. Subsequently, the statistical properties of the zero‐crossing rate and an appropriate detection criterion are discussed.