The central limit theorem describes how under certain conditions the distribution of the mean of a sample can be modelled by a normal distribution. This holds true for sufficiently large sample sizes, and allows for the use of normal distribution tools for inference on a population's mean.