During cooking, both temperature and time have a large effect on physical properties of meat, mainly due to protein denaturation. This work intends to evaluate a computer vision system to measure color changes and to develop a kinetic model to predict them during meat cooking. Pieces of semitendinosus muscle were heated by immersion in a thermostatic bath at constant temperature (from 40 to 100 ºC). Superficial color was measured using a computer vision system (CVS) and a colorimeter, and represented by means of the CIEL*a*b* color space. A kinetic model was developed to describe color changes during heating. The correlation coefficient between experimental and predicted values was 0.998 for colorimeter and 0.987 for CVS values. In addition, the dependence with the temperature of the fitting parameters was determined. The kinetic constants present a typical Arrhenius dependence. These results encourage coupling the kinetic model to a cooking model previously developed.