The demand for the consumption of milk and dairy products by the consumer market is very high. This makes it difficult to analyze the large number of milk samples for quality. In addition to the requirement to consider many quality attributes, there are usually large number of producers, who need daily milk evaluations. The aim of the study was to evaluate the efficiency of fuzzy logic in decision making for the classification of milk. In the fuzzification stage, physical and chemical characteristics of the milk were considered as input linguistic variables. For each linguistic variable, pertinence functions were created, and these were made considering the trapezoidal forms. In the inference stage, rules were established for the association of linguistic variables and output variables (adulterated, inadequate and adequate). To verify the efficiency of the modeled system, 1,000 adulterated, inadequate and adequate milk samples were computationally simulated. Precision was verified when automating decision making in the classification of milk by the fuzzy logic, totaling 100% of correctness. Therefore, the fuzzy system is an efficient tool for the classification of milk and can be used advantageously by professionals in the field in order to reduce human and financial resources.