Aimed at the problem that estimation accuracy reduces when square root cubature Kalman filter (SRCKF) faces measurement abnormalities and model mismatch in maneuvering target tracking, a robust adaptive algorithm based on feedback decision is proposed. The algorithm is to utilize the Huber function for processing the observation residual sequence to obtain a weight vector to correct the measurement covariance, and enhance the algorithm’s robustness to overcome measurement anomalies, and at the same time, multiple fading factors are introduced into the adjustable prediction error covariance, changing the filter gain, and enhancing the adaptability of the algorithm to solve the problem of model mismatch. Finally, according to the Mahalanobis distance, the abnormal error discrimination factor is constructed to realize the reasonable switching of the two processing methods by the feedback decision. As compared with the existing algorithms, the proposed algorithm can effectively deal with errors caused by measurement anomalies and model inaccuracy, and is good in the robustness and the adaptability.