Tapered steel members offer a better cross-section utilization along the member, which makes them an interesting and more economical alternative to prismatic ones. Yet, the design methodologies available do not provide a clear and sufficient guidance for the stability verification of such members. Alternatively, nowadays, the existing computer capacity and software programs provide an accessible and rapid means of reproducing the structural performance of members and systems, although they require beforehand validation to assure the plausibility of their predictions. For that, a full-scale experimental programme on non-uniform members was carried out, covering column, beam and beam-column tests. The test results are used to validate a numerical model commonly used for the assessment of stability design rules. In this paper, firstly, a global overview of the experimental tests is presented, which covers the test layout, member dimensions and the supplementary tests, essentially characterization of material properties, geometrical dimensions and imperfections, and residual stresses. The key results from each experiment are presented and discussed, they are further compared with numerical and analytical estimations of the member resistance. Finally, the experimental results provide physical validation of the design method proposed in Marques et al. (2012) for web-tapered columns. (C) 2018 Elsevier Ltd. All rights reserved.