This study was to optimize the fermentation process for protein enrichment of potato starch residue (PSR) by solid state fermentation with mixed strains. Three single-factor experiments were conducted to optimize the fermentation conditions, and a Plackett-Burman design followed by a central composite design of response surface methodology was investigated to optimize the medium constituents. The optimal fermentation conditions were fermentation temperature 28°C, fermentation period 72 h and solids content was 40%. A combination medium of 19.75 g potato starch residue, 6.38 g wheat bran, 2.86 g corn meal, 2.62 g urea, 0.15 g K2HPO4, 0.15 g MgSO4·7H2O, 0.075 g MnSO4·H2O and 0.015 g Fe2(SO4)3 was optimum for maximum protein
nitrogen content (3.75%) of fermented PSR, which was 3.4 fold over the raw material. The determination coefficient R2 (0.9889) and the adjusted determination coefficient R2(0.9790) indicated that the response equation provides a reasonable model for the central composite design experiment, and the lack of fit (P=0.0582) indicated that the model was adequate for prediction within
the range of variables employed. Due to the high-yield of protein nitrogen obtained and the low-cost nature of the optimal medium, this study indicates a possibility to establish production of protein feed using PSR as a basic substrate.