The subjective or behavioral value of travel time savings (SVTTS) calculated from discrete travel choice models as the trade off between cost and time in modal utility, represents the willingness to pay to diminish travel time (either in vehicle, waiting or walking) by one unit. This SVTTS can be shown to reflect the sum of at least two effects; first, the willingness to substitute travel time for other more pleasurable or useful activities and, second, the direct perception of the reduction of travel time itself. Regarding the first effect, one such substitute activity could be paid work, in which case the SVTTS will also include the additional money earned (or its equivalent goods consumption) in addition to the subjective value of work time. Hundreds of travel choice models have been estimated throughout the world from which values of time have been calculated. Its components, however, have never been estimated quantitatively. After Jara-Díaz (1998), in this article we take into consideration that travel (mode) choice and activity demand models come from a common microeconomic framework such that their specifications are linked. We show that estimating both type of models from the same population makes it possible to obtain all components of the SVTTS empirically (or to calculate them distinctly) because the models share some common parameters. This novel approach is experimentally applied using information on travel choices and home-work activities for two income groups collected in Santiago, Chile.