Eclipsing binary systems offer a unique opportunity to accurately measure the fundamental properties of stars. In the past years, masses and radii of low-mass stars in these systems have been obtained with uncertainties well below the 3% level, imposing stringent constrains on theoretical models. Several results revealed that there is a discrepancy between the radius predicted by models and that from the observations of about 5–10% depending on the mass range. On the other hand, effective temperatures are overestimated by models by about a 10%. These discrepancies have been usually attributed to stellar magnetic activity, however, the correct scenario is not yet clear. The progress in the understanding of stellar activity on low-mass stars and its relation with the radius discrepancies with models is summarized here both from the empirical and the theoretical points of view.