A computer code has been developed which simulates a Thomson scattering measurement, from the counting statistics of the input channels through the mathematical analysis of the data. The scattered and background signals in each of the wavelength channels are assumed to obey Poisson statistics, and the spectral data are fitted to a Gaussian curve using a nonlinear least‐squares fitting algorithm. This method goes beyond the usual calculation of the signal‐to‐noise ratio for the hardware and gives a quantitative measure of the effect of the noise on the final measurement. This method is applicable to Thomson scattering measurements in which the signal‐to‐noise ratio is low due to either low signal or high background. Thomson scattering data from the S‐1 spheromak have been compared to this simulation, and they have been found to be in good agreement. This code has proven to be useful in assessing the effects of counting statistics relative to shot‐to‐shot variability in producing the observed spread in the data. It was also useful for designing improvements for the S‐1 Thomson scattering system, and this method would be applicable to any measurement affected by counting statistics.