In that context, the study of binaries offers several advantages. First of all, the statistical properties of multiple systems bring important constraints on the models of formation and evolution. Depending on the mechanisms involved in the formation of these objects, the properties of the outcoming population, and in particular its properties of multiplicity, could be very different. Multiple systems can therefore be regarded as tracers of formation and evolution. Second, multiple systems provide a unique opportunity to study the physical properties of two (or more) objects of the same age which, since they most of the time have different masses, followed different evolutionary tracks. Both components of each binary are indeed expected to be coeval, thus removing part of the degeneracy in the mass-Luminosity (age-temperature) relation. Finally, the physical characteristics derived from the observations are still very much model dependent. By measuring dynamical masses and radii, independently from any models, binaries offer the opportunity to better understand the physical properties of this class of objects.