Evidence of the continuous rise of novel doping agents and novel doping strategies calls for the development of more accurate multi-target screening methods. Direct multi-target screening approaches are restricted to the targeted substances and their turnover. The development of effective "indirect" screening methods requires a priori a deep understanding of the metabolism of the substance. The biological passport has been demonstrated to be very effective, but it is limited to about 20 indirect parameters. The standard antidoping analytical methods are hence targeted and do not aim directly to identify unknown substances. Also, the detection of doping agents is limited by the excretion of the substance. This study considers metabolomics for the screening of performance enhancing hormone abuse by athletes, based on the following pieces of evidence: (1) hormones have a strong influence on human metabolism, changing several parameters in many tissues, organs, and bio-fluids; (2) metabolomics has been demonstrated to be a very accurate tool to depict the metabolic status of several organisms, tissues, and for several human diseases, including hormone deficiencies; (3) metabolomics has been demonstrated to be able to distinguish hormone-treated animals from controls in many species, without the need for a priori knowledge of the metabolism for the specific substance. The literature shows that metabolomics could be an appropriate tool to detect hormone abuse, keeping in mind the strength and the limitation of such an approach.