The deep hip muscles are often omitted in studies investigating hip contact forces using neuromusculoskeletal modelling methods. However, recent evidence indicates the deep hip muscles have potential to change the direction of hip contact force and could have relevance for hip contact loading estimates. Further, it is not known whether deep hip muscle excitation patterns can be accurately estimated using neuromusculoskeletal modelling or require measurement (through invasive and time-consuming methods) to inform models used to estimate hip contact forces. We calculated hip contact forces during walking, squatting, and squat-jumping for 17 participants using electromyography (EMG)-informed neuromusculoskeletal modelling with (informed) and without (synthesized) intramuscular EMG for the deep hip muscles (piriformis, obturator internus, quadratus femoris). Hip contact force magnitude and direction, calculated as the angle between hip contact force and vector from femoral head to acetabular center, were compared between configurations using a paired t-test deployed through statistical parametric mapping (P < 0.05). Additionally, root mean square error, correlation coefficients (R2), and timing accuracy between measured and modelled deep hip muscle excitation patterns were computed. No significant between-configuration differences in hip contact force magnitude or direction were found for any task. However, the synthesized method poorly predicted (R2-range 0.02–0.3) deep hip muscle excitation patterns for all tasks. Consequently, intramuscular EMG of the deep hip muscles may be unnecessary when estimating hip contact force magnitude or direction using EMG-informed neuromusculoskeletal modelling, though is likely essential for investigations of deep hip muscle function.