The classical approach to detect the abuse of banned substances, in which biological samples are semi-quantitatively tested for the presence of a banned substance or a marker metabolite, cannot be directly applied in case a natural background level can be present. However, even when this background level is relatively low, the mere fact that the presence of a particular analyte due to general or specific physiological conditions cannot be excluded with certainty, asks for a different analytical testing strategy. This reflection paper summarizes the approaches a wide range of authors followed when trying to solve this challenge. The classical testing model comprises two steps: a screening procedure focussing on selecting suspect samples while avoiding false negative results, followed by a confirmatory step focussing on the unambiguous confirmation of the identity of the analyte. Screening for the abuse of (semi) natural hormones can follow a similar approach since also, in this case, the outcome of the screening procedure is either a result “compliant”, no indication of treatment with a banned substance, or “suspect”, treatment cannot be excluded. However, the definition of a suspect is slightly different in this case. In the classical model, suspect means that there are indications that the target analyte is present in the sample tested. When testing for natural hormones, suspect means that there are indications that the animal from which the sample was taken was treated with a banned substance. This difference has a significant impact on the confirmatory procedure to be applied. Three different types of confirmatory methods for (semi-)natural compounds can be identified. (1) Confirmation of the identity of a compound of which no endogenous source is known. This type of confirmatory method is similar to the approach used for exogenous compounds, e.g. synthetic esters of (semi-)natural compounds or confirmation of a biomarker proven not to be present endogenously. (2) Confirmation based on statistical solid differences between the analyte isolated from a sample and an endogenous reference compound. This situation applies in case combustion-Isotope Ratio Mass Spectrometry is used for confirmation. In selected cases, the difference in C12/C13 between endogenous and synthetic forms of a molecule can be such that discrimination is possible with a significant solid difference. (3) Confirmation on the basis of another significant difference between data obtained for a particular animal or group of animals and the corresponding reference population. This could involve exceeding threshold levels, differences in ratio’s between selected compounds and differences in profiles of metabolites and/or biomarkers. Until now, this third type of confirmation is usually considered for screening purposes only, and this reflection paper is presented as such. Depending on the significance of the differences observed, use as a confirmatory method can be considered. Currently there are no criteria available for this type of analysis. However, for this, scientific and statistical considerations are not enough. Full agreement with Competent Authorities will be needed. Specific recommendations are prepared and presented in this manuscript for the different groups of (semi-) natural hormones.