Analyses of vitamin supplements A and E in food samples are performed mostly with high performance liquid chromatography (HPLC). In majority of cases, sample preparation preceding HPLC implies saponification, a step critical to heat sensitivity of analytes. The method of saponification is clearly defined by ISO standards, however, two important factors, temperature and time of saponification are only given in value ranges instead of exact settings. Resolving this deficiency with the promise of eliminating time and cost consuming experimental probes, statistical experimental design (SED) is introduced to find optimum settings of temperature and time for the best recovery of vitamin supplements in food samples. Finding the optimum settings in SED was supported with Statsoft Statistica 13 For illustrating SED, margarine samples supplemented with vitamin A and E were applied.