In the context of the control of the illegal administration of natural steroid hormones in cattle husbandry, an attempt was made to establish the decision levels for sex steroid hormones in the plasma of adult cattle, taking into account the effect of the treatment. Bulls and heifers were treated with two injections, at a two week interval, of an estradiol–testosterone cocktail. Steroid hormone and biochemical precursor concentrations were measured in plasma samples by using specific radioimmunoassays, before and after the treatment. When the treatment significantly (p < 0.05) modified a hormone concentration, a decision level was established for that hormone concentration. At each decision level, a score was assigned that represented the percentage of treated animals detected when the decision limit was applied. For heifers, 17β-estradiol and testosterone concentrations in plasma, which increased after the treatment, are the best criteria to use to detect treated animals, with decision limits of 20 pg ml^–1 and 125 pg ml^–1, respectively. In the instance of bulls, both testosterone and steroid biochemical precursor concentrations decreased in the plasma after the treatment. We proposed decision limits of 1500 pg ml^–1 and 28 pg ml^–1 for testosterone and androstenedione concentrations, respectively, the bulls displaying concentrations below these limits being positive. We observed that the repetition of the injection increased the score of the decision limit. The scores for testosterone are 70%, 14 d after the first injection and 100% 14 d after the second injection, and for androstenedione, these scores are 60 and 100%, respectively.