Injection-moulded (IM) short-glass fibre reinforced thermoplastics (SFRPs) are light-weight, have short production cycles, and can be used to manufacture complex 3D-shaped components. However, existing failure criteria for SFRPs (e.g. based on the Tsai-Hill criterion) underestimate the failure load of components, because they consider failure initiation only. The objective of this work is therefore to develop an FE methodology to predict failure of IM-SFRPs, accounting for progressive failure. Polyamide 6.6 reinforced by 50% (wt) glass-fibres (PA66-GF50) from Asahi Kasei Corporation was used to injection-mould the following samples: (i) dog-bone specimens and Compact-Tension (CT) specimens (machined at several angles from 120 × 80 × 2 mm plates, results shown in Figure 1(a)), and (ii) subcomponents representing an automotive engine mount. These samples were tested under 23°C and 80°C, and under Dry As Moulded (DAM) or 50% of Relative Humidity (RH50). The field of fibre orientation tensors in the subcomponents (generated due to the injection-moulding process) was predicted using Moldflow software, and mapped onto an Abaqus Finite Element (FE) model of the subcomponent test, using Digimat-MAP. Digimat’s elastic-plastic material model was calibrated with the dog-bone tensile testing data, accounting for the anisotropy and material nonlinear-response. Figure 1 (b) shows load-displacement curves obtained experimentally and from the coupled simulations, using two approaches to predict failure: • The conventional approach (“TH”), corresponding to failure initiation as predicted by the Tsai-Hill criterion. While there is good agreement on the initial stiffness and progressive softening, this approach predicts failure about 20% earlier than observed in the tests. • Our proposed FE methdology, accounting for the fracture toughness of the material through cohesive zone modelling (“CZM”). This shows excellent agreement between the predicted failure point and the test results, within 3% error for all environmental conditions tested. This work demonstrates that accounting for the finite fracture toughness of IM-SFRPs is required to overcome the limitations of conventional methodologies, and accurately predict the ultimate failure of components and how they are affected by environmental conditions.