Background <p>Analysis of finite elastic–plastic deformation requires a good constitutive description of the inelastic response. A key element is the hardening function, typically evaluated from nominal data obtained from a uniaxial tension test using the Bridgman correction method in the post-necking regime. However, the true stress–strain data obtained in this way depend crucially on the quality of the optical measurements. Based on modern and sophisticated technology, several new and demanding methods have emerged over the last decade to improve the accuracy of the Bridgman correction method.</p> Objective <p>To develop and validate an accurate method, combining standard experimental and numerical methods for the evaluation of the true stress–strain response of an elastic–plastic material.</p> Method <p>Experimental force-displacement data for a gauge section of a round bar specimen are used as input to an inverse finite interval (IFI) method to construct the true stress-strain response at discrete points by minimising the relative error between finite element simulations and experimental results.</p> Results <p>The inverse finite interval method is validated both numerically and experimentally. Synthetic force–displacement data are generated numerically for a set of power-law hardening materials, and the proposed method is then applied to reconstruct the true stress–strain curve. The method is proven to be robust, and reconstructed true stress–strain curves converge to the exact power-law function. Three markedly different steel grades are used for experimental validation by comparing the deformed geometry in the necking region between experimental and numerical results, showing good agreement.</p> Conclusion <p>The proposed inverse finite interval (IFI) method is an expedient and accurate method for evaluation of the true stress–strain curve from nominal data extracted from a uniaxial tension test.</p>