A validated mathematical model has been used to perform multi-objective optimization of beef roasting, considering the minimization of both cooking time and weight loss. Simulations were performed using irregular geometrics models of beef semitendinosus muscle and constant oven temperature. Minimum temperature of 72 degrees C at the coldest point was imposed as a constraint. From model results a compromise situation was encountered, since no operative condition lead simultaneously to optimal values of both objectives, i.e., an increase in oven temperature leads to a decrease in cooking time, between -0.25 and -0.325 min/degrees C, and an increase in weight loss, between 0.175 and 0.275%/degrees C. In this sense, not one, but a set of optimal solutions was found in the Pareto sense. Experimental cooking tests were performed, which are in good agreement with model simulations. Furthermore, energy consumption for each optimal solution obtained in the optimization problem was estimated, being lower using low oven temperatures. (c) 2012 Elsevier Ltd. All rights reserved.