Stellar evolutionary models simulate well binary stars when individual stellar mass and system metallicity are known. The mass can be derived directly from observations only in the case of multiple stellar systems, mainly binaries. Yet, the number of such stars for which accurate stellar masses are available is rather small. The main goal of this project is to provide realistic mass estimates for a homogeneous sample of about a thousand FGK single stars, using four different methods and techniques. We present the masses inferred according to each one of these methods as well as a final mass estimate consisting in the median of the four mass estimates. The procedures evaluated here include the use of stellar evolutionary models, mass-luminosity relation and surface gravity spectroscopic observations. By combining the results obtained with different methods, we determine the best mass value for each individual star, as well as the associated error budget. Our results confirm the expected consistency between the different mass estimation methods. None the less, for masses above 1.2M_{sun}_, the spectroscopic surface gravities seem to overestimate the mass. This result may be a consequence of the spectroscopic surface gravities used in this analysis. Nevertheless, this problem is minimized by the fact that we have several approaches available for deriving stellar masses. Moreover, we suggest an empirical procedure to overcome this issue.