The success of the thixoforming process depends on the possibility to confer to material, when it is found in the semisolid state, a microstructure characterized by globular particles of solid phase surrounded by a continuous film of liquid phase; such microstructure is obtainable through particular thermo-mechanical treatments. In the present research, in order to optimize the influence of process parameters in the step in which the thixotropic properties are conferred to the AA7075 aluminum alloy, the statistic technique of the Design Of Experiments (DOE) has been used. The advantages in the application of such technique are expressible in terms of reduction the times of development of process and more efficient use of resources. A Central Composite Design factorial plan with two levels has been realized to allow the evaluation of experimental error and to check the adequacy of the model. The experimental tests foreseen from the same plan have been therefore performed. Using the method of the response surfaces (RSM), the function of response has been formulated, and the analysis of the experimental data has been realized by the linear regression method. Finally, the analysis of variance (ANOVA) allowed to value the causes of variability of the results. Statistic test has been performed on the significance of single factors and their interactions.