Within the EC-financed project "Feasibility Study for the Production of Certified Calibrants for the Determination of Deoxynivalenol and other B-Trichothecenes", an intercomparison study was performed with 13 European participants.Main goals of the intercomparison study were to check the feasibility of a small batch of gravimetrically prepared calibrants, to directly compare common and individually prepared calibrants, to test the practicability of toxin mixtures as calibrant solutions and finally to give recommendations for the means of certification. Additionally, it focused on the comparison of gas chromatography (GC) and high performance liquid chromatography (HPLC) for the determination of pure type-B trichothecene solutions, which is described in this publication.The participating laboratories received calibrant solutions as well as toxin solutions of unknown concentration and employed mainly HPLC-UV; GC-ECD (electron capture detection) and GC-MS (mass spectrometry) methods were used less often.The intercomparison study generally suffered from a high rate of outliers (22% of all the data). Throughout the study, 48% of all GC results were classified as outliers and it soon became apparent, that GC results highly infuenced the outcome of the study and that the used GC methods were not robust enough for the certification of type-B trichothecene calibrants. The high discrepancy between HPLC and GC results in the intercomparison study presumably lies in the crucial step of derivatisation.