The extraction of anthocyanin from Bokbunja (Rubus coreanus Miq.) marc generated during traditional wine processing was optimized using response surface methodology (RSM). A face-centered cube design (FCD) consisting of 17 experimental runs, including five replicates at the center point, was used to investigate the effects of the three variables (solid-liquid ratio, time, and temperature) on anthocyanin extraction, and the results showed that the relationship between the three variables and the total anthocyanin content followed a quadratic model (R2=0.8853). In addition, the RSM analysis predicted that the optimum conditions for extraction consisted of a solid-liquid ratio of 20, a time of 60min, and a temperature of 60 degrees C. Verification tests performed under these optimum conditions gave 34.7+/-1.4mg/100g of anthocyanin, which was close to predicted value of 37.2mg/100g. Additionally, analysis of water extracts prepared using the predicted optimum conditions revealed that the carbohydrates (sugar and pectin) in Bokbunja marc underwent significant variation toward the formation of by-products (glycerol and uronic acids) during yeast fermentation, and that the amount of anthocyanin produced was reduced 10-fold when compared to the original extraction. Further, the results of HPLC-PDA-MS/MS analysis of the anthocyanins extracted from Bokbunja marc revealed the presence of six anthocyanin components, which were tentatively identified as cyanidin 3-O-sambubioside, cyanidin 3-O-xylosylrutinoside, cyanidin 3-O-rutinoside, pelargonidin 3-O-rutinoside, delphinidin 3-O-rutinoside-?, and delphinidin 3-O-glucuronide.