The significance of the models generated for each response was tested by analysis of variance (ANOVA) with a significance level of 5 for all tests. The representation of the mixture system was constructed using a triangular graph with level curves for each dependent variable based on the fitted model. The polynomial models that fit the responses of optimum cooking time, loss of solids, mass increased and chroma b ere chosen as the best fresh pasta formulations based on the mixture components (DPP, EPP and AF) to obtain the most desirable spaghetti. The positive char- acteristics of the samples using the desirability test consisted of the following: a short optimum cooking time, small loss of solids, in termediate mass increase, and low chromaticity coordinate Statistica 70 software (Statsoft, b Statistica 70 for Windows, Tulsa, USA) was used to determine the experimental design. perform the statistical analysis and construct the graphs.