Abstract: Using edible cactus as raw material,the response surface-principal component analysis method was used to study the cactus fermented wine process. By studying the effects of material to water ratio,yeast inoculum,initial sugar content and fermentation temperature on the quality of cactus fermented wine,the optimum process was determined. The results showed that the regression coefficient of the two polynomial regression model established by principal component analysis was significant,and had better fit degree (p<0.0001,R²=0.9599). The best process conditions for cactus fermented wine were:material water ratio of 1:4,yeast inoculum of 0.05%,initial sugar content of 25%,fermentation temperature of 20℃. Under this condition,the cactus fermented wine had an alcohol content of 12.9%,flavonoid content of 6.2 μg/mL,total phenol of 38.6 μg/mL,total acid of 4.9 g/L,total sugar of 5.2 mg/mL,and the standardized comprehensive score was 0.899,which was close to the theoretical standardized comprehensive score of 0.916.