A study was conducted to develop a solvent system that will clean egg yolk samples and concentrate cholesterol oxides effectively before analysis. Cholesterol oxide standards or lipid samples (0.2 g) loaded onto a silicic column were washed with a portion of Solvent I (hexane/diethyl ether, 9:1, vol/vol) and then with Solvent II. Four different Solvent II preparations (Solvent IIa, hexane:ethyl acetate = 4:1; Solvent IIb, hexane:ethyl acetate = 1:1; Solvent IIc, hexane:ethyl acetate:diethyl ether = 2:1:1; Solvent IId, hexane:ethyl acetate:diethyl ether = 4:1:2, vol/vol/vol) were prepared and the purification efficiencies of Solvent II solutions for neutral lipids, cholesterol, and phospholipids in the column were compared. Yield study using cholesterol oxide standards showed that one or more of the cholesterol oxide standards were eluted by the Solvent IIb and Solvent IIc, but Solvent IIa and Solvent IId did not elute any of the cholesterol oxides during washing. Egg samples prepared with Solvent IIa showed greater amount of cholesterol oxides than those prepared with Solvent IId, probably due to incomplete purifying of phospholipids and interference. However, the amounts of cholesterol oxides in cooked meat prepared with the two purification solvents were not different. Because egg yolk contains very large amounts of phospholipids and cholesterol compared with other foods, at least twice as much Solvent IIa as Solvent IId was required to properly clean egg yolk samples. It was concluded that purification solvents should be selected by sample types, and Solvent IId (hexane:ethyl acetate:diethyl ether = 4:1: 2) was superior to Solvent IIa (hexane:ethyl acetate = 4: 1) for egg yolk samples.