 The paper reviews and classifies statistical transformations for post-processing regional climate model, RCM, output into three categories, distribution derived, parametric and non-parametric, each with different underlying assumptions. A real-world application using Norwegian precipitation station observations shows that non-parametric transformations have the highest skill in reducing biases in RCM precipitation. This article was authored by J. E. Haubin, T. Enginskogen, J. B. Brands and others.