The use of thyreostatic drugs, like thiouracil (TU), in animal production has been banned for over three decades by the European Union, due to potential teratogenic and carcinogenic effects of its residues upon human consumption. Besides, thyreostats induce water retention in livestock, causing fallacious weight gain and deterioration of meat quality. Development of more competent analytical methods gave rise to sporadic TU detection in urine of untreated animals, questioning the actual synthetic origin TU. Research showed that TU can be formed upon digestion of Brassicaceae feeds in vivo and in vitro, which called for a means of differentiation between endogenous formation of TU and illicit administration. Therefore, in the present study, a routinely applicable liquid chromatography (LC) ion trap multiple mass spectrometry (MS(2)) method for TU analysis in animal feeds was optimised and validated, according to CD 2002/657/EC. A fractional factorial Plackett-Burman design was used to optimise the extraction procedure for TU from Brassicaceae and non-Brassicaceae feeds. This resulted in the discrimination of five influential factors (amount of feed, myrosinase, pH 7 buffer, 3-iodobenzyl bromide and elution solvent), for which the most optimal conditions were perfected. The limit of quantification for TU amounted 0.5 ng g(-1). The individual recoveries for TU ranged between 90.9 and 99.7%. Good results for repeatability and intra-laboratory reproducibility (RSD%) were observed, i.e. ≤6.0 and ≤5.2%, respectively, for TU. Excellent linearity was proven based on determination coefficient (R(2) ≥ 0.99) and lack-of-fit test (F test, α = 0.05). Subsequently, a selection of feeds sampled during European national monitoring campaigns were evaluated with the present method showing concentrations ranging from 0.32 to 20.60 ng g(-1), demonstrating the relevance of the method in the analysis of TU from animal feeds.