Abstract: A new general two-stage algorithm is originally proposed to reduce the computational effort for maneuvering target tracking in mixed coordinates. The augmented state Kalman estimators, which are based on the jerk modeling, are computationally expensive. The conventional input estimation techniques assume constant acceleration level and there are not covered a generalized input modeling. In this research, an innovative scheme is developed to overcome these drawbacks by using a reduced state Kalman estimator with a new structure, which is optimal for general conditions. In addition, the proposed scheme is an unbiased filtering algorithm applied in mixed coordinates based on the pseudo linear measurements.