Response surface methodology (RSM) was employed to investi-gate the variations in recovery rates with respect to the preparationof conditions including extraction solvent volume, the amountsof sodium acetate, PSA, and C18. The optimal composition of thefour variables was determined by using a central composite design