AbstractApplication of statistical testing to mixing.Assessment of the homogeneity of mixtures of solids and determination of the necessary mixing times are performed with the aid of statistical testing methods. The results from the basis for necessary decisions. Since reliable statements are fundamentally impossible in mathematical statistics, all such decisions are, with a certain probability, false, i.e. involve some risk. In studies of mixing, the first kind of error concerns the risk of rejecting a mixture although its homogeneity satisfies requirements, while the second kind of error concerns the risk of accepting a mixture although its homogeneity is insufficient. It is shown that barriers can be introduced for both kinds of error, and how to do this. Both the sample size and the limits of the acceptance range for the sample variance are thus laid down. If no decision is to be made concerning acceptance or rejection of a given mixture, but information is merely required about the mixing quality after a given time then this is accomplished by statement of confidence intervals.