As an alternative of single filler system, hybrid fillers system is beneficial for the fabrication of elastomer‐based piezoresistive strain sensors with high strain sensitivity for structural health monitoring and human motion detection. However, a comprehensive understanding for the effect of the secondary filler dimensionality on piezoresistive strain sensor sensitivity by hybrid carbon nanofillers strategy is missing. Here, we report a comparative study about the effect of branched carbon nanostructures (CNS, three‐dimensional, 3D) combined with carbon black (CB, 0D) or carbon nanotubes (CNT, 1D) on the piezoresistive behavior of thermoplastic polyurethane (TPU)‐based composites. At either the same total filler content or similar electrical resistivity, using CNS‐CB hybrid is more efficient than using CNS alone or CNS‐CNT hybrid for improving piezoresistive sensitivity of the TPU composites. For instance, the composite 1.5CNS‐1CB has a gauge factor (GF) of 21 and 99 at strain ε = 50 and 100%, respectively, while the composites 2.5CNS and 1.5CNS‐1CNT have GF values of 4.3 and 9.1 at strain ε = 50% and 14 and 20 at strain ε = 100%, respectively. Besides, the TPU composite with CNS‐CB shows a larger response to finger bending (50° on an index finger) by comparison with that contains only CNS, indicating potential applications for human motion detection. This study introduces a facile way to fabricate strain‐sensitive sensors aimed at strain of 50%–100%, strengthening understanding the selection of hybrid carbon nanofillers with different dimensionalities to optimize piezoresistive sensor design.