SummaryThe glyceride composition of soybean oil has been calculated according to even and random distribution patterns and has been compared with experimental fractionations obtained by adsorption analysis, liquid‐liquid extraction, and crystallization. Evidence is presented which indicates that soybean oil does not follow the elementary conception of even distribution. Fractions have been isolated by these three methods with iodine values lower than that permitted under the rule of even distribution. Further data indicate the occurrence of di‐saturates and dilinolenin, neither of which is permitted under the rule of even distribution.The ultimate proof of the distribution pattern existing in soybean oil, whether random or partial random, must await the development and application of more efficient methods of fractionation. At the present time separation of soybean triglycerides appears limited by the efficiency of fractionation procedures rather than by the pattern of fat acid distribution.