The characteristic function method is employed to study the output spectral density and signal-to-noise ratio of a full-wave error function limiter. The input signal and noise are considered to be sample functions from independent Gaussian random processes. The results are obtained as a function of the input signal-to-noise ratio, degree of limiting and bias, Limiter and signal parameters are expressed in terms of known functions which may be easily programmed to yield numerical results. †Communicated by the Authors.