In order to establish a comprehensive food quality evaluation method, 2 mathematical modeling evaluation methods were applied (grey system and entropy weight method). The experimental material was Flammulina velutipes feet (FVF). The by-product of F. velutipes was divided into 3 equal portions for frying, and the resulting fried FVF (FFVF) were evaluated. Based on the ranking of 50 people’s food attribute preferences, grey relational analysis (grey system) was used for subjective evaluation. The results showed that people’s preference ranking for the 3 parts was: lower > middle > upper. The entropy weight method was then used to verify the results based on the FFVF nutritional indexes (total sugars, protein and dietary fiber) and flavor indexes (flavor amino acids, flavor 5’-nucleotides and electronic tongue flavor score). The validation results were lower > middle > upper, which was consistent with the subjective evaluation results, suggesting the lower part of FVF was more suitable for the frying process. In conclusion, values of different parts of FFVF were rated, which provided a new processing method for the utilization of F. velutipes by-products. It also provided a novel means for the evaluation of processed products in food engineering.