The dual mode INS high performance system using an IMM mixing method presented in this disclosure utilizes a MEMS-based IMU as the inertial sensor due, in part, to low cost and low power consumption. A dual filtering system, one with fifteen state EKF and the other with twenty one state EKF is mixed in real-time to address dynamic overlapping and continuous transition from a low dynamic mode to a high dynamic mode. The IMM estimation scheme is employed to compute the real-time mixing coefficients via the IMM probability mixing formulation. The present disclosure is applicable to earth to orbit platforms, spinning projectiles, and the like to successfully reconstruct the object's state estimate vector even under a high angular rate and high g operating environment.