This paper describes a statistical criterion for evaluation and selection of different testing methods for solid biofuels taking into consideration accuracy, precision, sensitivity, reproducibility, repeatability, testing costs and testing time. The signal-to-noise ratio as suggested by Taguchi has been used in a way similar to a traditional method (analysis of variance, ANOVA) used for this purpose. Some simulated examples are described to illustrate the development of the proposed technique. Application to real situations can be made by treating experimental data in a similar way.