A method was developed to determine acacetin, luteolin and quercetin in Chrysanthem indicum L by β-CD modifiedcapillary electrophoresis. Through the investigation of the effects of buffer pH and concentration, applied voltage, SDS concentration andβ-CD, the analytical conditions were optimized. Under the optimized conditions, the two analytes were well separated in 8 min. A goodlinear relationship between the peak area and concentration was found in the 2-400μg/mL,2-500 μg/mL and 2-500μg/mLconcentration range for acacetin, luteolin and quercetin, respectively. The relative standard deviation based on migration time and peakarea were 0.51%, 0.76%, 0.83% and 2.17%, 2.42%, 2.68% for acacetin, luteolin and quercetin, respectively. The detection limits basedon three time noise were 0.6μg/mL, 0.5μg/mL and 0.6μg/mL for acacetin, luteolin and quercetin, respectively. The method wasverified by real sample analysis and running standard addition and recovery experiments with satisfactory results. A method was developed to determine acacetin, luteolin and quercetin in Chrysanthem indicum L by β-CD modified capillary electrophoresis. Through the investigation of the effects of buffer pH and concentration, applied voltage, SDS concentration and β-CD, the analytical conditions were optimized. Under the optimized conditions, the two analytes were well separated in 8 min. A good linear relationship between the peak area and concentration was found in 2-400 μg / mL, 2-500 μg / mL and 2-500 μg / mL concentration range for acacetin, luteolin and quercetin, respectively. The relative standard deviation based on migration time and peakarea were 0.51%, 0.76%, 0.83% and 2.17%, 2.42%, 2.68% for acacetin, luteolin and quercetin, respectively. The detection limits based on three time noise were 0.6 μg / mL, 0.5 μg / mL and 0.6 μg / mL for acacetin, luteolin and quercetin, respectively. The method was verified by real sample analysis and running standard addition and recovery experiments with satisfactory results results.