Consumers are now increasingly aware of the impact of meat production on animal welfare and the environment.Simultaneously, there has been a decline in meat consumption and a demand for high-quality meat (in terms of sensoryas well as nutritional quality). This study aims to propose a methodological approach that uses breeding practices toestimate meat quality, aiming to achieve optimal quality and meet consumer demand. To achieve this goal, we havedeveloped an updated version of NSGA-II (Non-dominated Sorting Genetic Algorithm II). This algorithm generatesa set of candidate solutions, selects the best individuals based on their fitness, and applies genetic operators such ascrossover and mutation to generate new offspring. The decision space is defined by the variables X related to themanagement of breeding practices, while the objective space Y represents the variables related to the sensory and/or nutritional quality of the meat to optimize. To ensure accuracy and precision, the fitness value of each objective isassessed using a multiple linear regression model. An AIC (Akaike Information Criterion) approach is then mobilizedto select the most relevant model for each objective. Once a new population is evaluated using the selected models,the Pareto front approach is utilized to identify the non-dominant variables in the multi-objective space. In order toprevent the algorithm from getting trapped in local maximum scenarios, a crowding distance method is employedto maintain population variability and to ultimately reach the global maximum. With this approach, we can generatethe best breeding practices for each breed/type of animal and optimize quality. Using the hypervolume approach,we can compare the different optimum front scenarios and recommend, for example, the best breed according to theobjectives. In conclusion, this study presents an updated methodological approach for estimating meat quality usingbreeding practices, which has the potential to improve meat quality and meet consumer demands.< Réduire