The interpretation of Weibull probability plots of mechanical testing data from castings was discussed in Part 1 (M. Tiryakioglu, J. Campbell: Metall. Mater. Trans. A, 41 (2010) 3121-3129). In Part II, details about the mathematical models of Weibull mixtures are introduced. The links between the occurrence of Weibull mixtures and casting process parameters are discussed. Worked examples are introduced in five case studies in which six datasets from the literature were reanalyzed. Results show that tensile and fatigue life data should be interpreted differently. In tensile data, Weibull mixtures are due to two distinct defect distributions, namely “old” and “young” bifilms, which are a result of prior processing and mold filling, respectively. “Old” bifilms are the predominant defect and result in the lower distribution, whereas “young” bifilms results on the upper distribution. In fatigue life data, Weibull mixtures are due to two failure mechanisms being active: failure due to cracks initiating from surface defects and interior defects. Surface defects are predominant and interior defects lead to fatigue failure only when there are no cracks initiated by surface defects. In all cases, only the mutually exclusive Weibull mixture model was found to be applicable.