The term Nyquist frequency has two related meanings, either as a measure of a digital system's performance or of an analogue signal's requirements for a digital system. In the strictest sense it is a characteristic of processing systems that defines the maximum input frequency the system can resolve. Any frequency components in the input that are higher than this frequency result in an incorrect digital representation of the signal. A slightly different use of the term is as a characteristic of signals rather than systems. More properly called the Nyquist rate, this defines how frequently a digitizer must sample a specific analogue signal to avoid this same distortion.

Perform the steps in this section if the goal is to find an existing digital system's Nyquist frequency, which describes the performance of the system in terms of required input signal characteristics.

Find the sample rate of the digital system from its specifications. Sample rate is usually expressed as samples per second, but it is also equal to the digital symbol rate, or baud rate, of the system which is in symbols per second.

Divide this sample rate or baud rate by two. The result is the system's Nyquist frequency, the maximum analogue input frequency the digital system can correctly process.

Perform these steps to find the Nyquist frequency of a specific signal, more properly called the Nyquist rate. It specifies the performance required of a digital sampler in providing a proper representation of this signal.

Find the maximum frequency component of the signal. If it is specified, skip to step 5. If not, move to the next step.

If possible, determine the upper limit of the signal frequency from the specifications for the circuit elements through which the signal passes. This entails finding the minimum listed bandwidth of filters or other devices that limit the signal's upper frequency. If this information is available, skip to step 5. If not, move to the next step.

Measure the signal's upper frequency limit using a broadband frequency domain instrument like a spectrum analyzer. Follow the instructions for the instrument in making this measurement.

Multiply the signal's upper frequency by two. The result is the Nyquist rate for this specific signal, sometimes called the Nyquist frequency. It is the required sample rate of any digital system necessary to avoid distortion of this signal when converting it to a digital representation.