Finite element analysis (FEA) is the leading numerical technique for studying joint biomechanics related to the onset and progression of osteoarthritis. However, subject-specific FEA of joint mechanics is a time- and compute-intensive process limiting its clinical applicability. We introduce and evaluate a novel hybrid modelling framework combining discrete element analysis (DEA) and FEA for computationally efficient evaluation of cartilage mechanics in the hip joint. In our approach, the hip joint contact mechanics are first estimated using DEA and subsequently used as input for matching FEA models, substantially reducing model complexity. The cartilage mechanical responses obtained using the hybrid DEA-FEA method were evaluated for subject-specific hip joint geometries from five asymptomatic individuals under loading conditions typical to normal walking gait and compared to conventional FEA in terms of peak intra-tissue mechanical stresses and model run-times. The hybrid DEA-FEA method had a median run-time of 3.6 min per subject (64-core processor, 512 GB RAM) and produced minimum principal (compressive) stress estimates comparable to stresses obtained using conventional FEA models with a median run-time of 96.2 min. On average, the peak compressive stresses obtained using the hybrid DEA-FEA approach were 0.06 MPa (95 % confidence interval: -0.86-0.99) lower than the stresses estimated with conventional FEA. Despite up to 1.4 MPa differences at individual gait time-points, the results indicate that the proposed hybrid DEA-FEA method enables estimation of hip cartilage mechanics in a fraction of time compared to conventional FEA, facilitating implementation in large cohort studies and clinical applications.