The relationships between mass and size and other parameters of raisin berries are needed for grading or classifying purposes. In this study, the mass of raisin berries was measured and predicted by applying different physical characteristics with 12 linear and nonlinear models in two different classifications. It is also possible to statistically characterise raisin berries on the basis of dimension for classifying and separating targets. For this reason, the Weibull distribution model was used to describe the process under study. The results showed that from the economic and agronomical point of view, a suitable grading system of raisin berries mass could be selected based on the major diameter as the nonlinear relation: M = 0.068 e0.157L, R2 = 0.769. The information on statistical or physical modelling of raisin berries can be alternatively or synthetically utilized for classifying or grading targets.