The study investigates the optimal extraction process of total flavonoids from Sophora davidii(Franch.)Skeels with single factor experiments,response surface methodology(RSM)analysis and genetic algorithm-neural network(GA-ANN).The extraction time,solid-liquid ratio and ethanol concentration are the variable factors,and the extraction rate of total flavonoids is the response index.The optimal results of response surface experiment are as follows:when ethanol concentration is 60%,solid-liquid ratio is 1∶30(g/mL)and extraction time is 90 min,the extraction rate of total flavonoids is 0.86%.Through using response surface data as learning data,the optimal experimental conditions obtained by genetic optimization using artificial neural network are as follows:solid-liquid ratio,ethanol concentration and extraction time are 1∶20,80%and 2.22 h,respectively.Under these conditions,the average extraction rate of total flavonoids is 0.77%.The results show that the response surface optimization technique is more effective for the extraction of total flavonoids from wild wolfsthorn.