Objective: To optimize extraction process of Bauhinia championi. Method: Taking ethanol concentration

ratio of material to liquid and reflux time as independent variables

overall desirability of dry extract yield

contents of 3'

4'

5'

5

7-pentamethoxyflavone(flavonoids 1) and 3'

4'-methylenedioxy-5'

5

6

7-tetramethoxyflavone(flavonoids 2) as dependent variable

data of overall desirabilities were fitted to a second-order polynomial equation and regressed by multiple linear

response surface methodology was used to optimize extraction technology and predictive analyze.HPLC was adopted to determine contents of flavonoids 1 and flavonoids 2 with mobile phase of acetonitrile(A)-water(B) gradient elution(0-20 min

30%-40% A;20-36 min

40%-48% A) and detection wavelength of 324 nm. Result: Complex correlation coefficient of binomial fitting was higher with overall desirability of 0.995 6;Optimal extraction technology was as following:extracted 2 times with 14-fold the amount of 45% ethanol

116 min per time;The mass fractions of flavonoids 1 and flavonoids 2 were 0.046% and 0.027%

dry extract yield was 21.21%. Conclusion: Central composite design-response surface methodology was convenient and highly predictive for optimizing extraction technology of B. championi.