Abstract A nine-factor central composite design was used for studying the effects of nine major processing variables on the quality of frozen surimi and the optimal surimi processing system. The results of analysis of variance showed that the total effects of the leaching water (L), leaching cycle (N), amount of salt (S), amount of sugar (U) and frozen storage (F) on the gel quality and whiteness of the product were very significant but total effects of setting and heating temperature and time (W, X, Y and Z) were insignificant. Canonical analysis indicated that the stationary points were saddle points for both of the response surfaces of gel strength and whiteness. Ridge analysis indicated that maximum gel strength will result from the use of a large amount of leaching water (L = 5.82), a small number of leaching cycles (N = 0.11), a high salt content (S = 3.03%) and medium sugar, frozen storage, setting and heating conditions (U = 3.3%, F = 3.4 months, W = 36.0 °C, X = 71.5 min, Y = 68.0 °C, Z = 8.12 min). Minimum whiteness will result from short frozen storage times (F = 0.03 month) and medium leaching, grinding, setting and heating conditions (L = 4.08, N = 2.91, S = 1.83%, U = 4.02%, W = 35.0 °C, X = 64.7 min, Y = 70.9 °C, Z = 7.97 min). In further investigations, the setting and heating conditions might be best fixed at moderate values and the effects of L, N, S, U and F should be concentrated. The lack-of-fit for the model was highly significant, which indicated that the quadratic model did not fit the data very well. Further research is needed to optimize the surimi processing system.