In this paper, a new hybrid unscented Kalman (UKF) and unscented [Formula: see text](U[Formula: see text]F) filter is presented that can adaptively adjust its performance better than that of either UKF and/or U[Formula: see text], accordingly. In this way, two Takagi-Sugeno-Kang (TSK) fuzzy logic systems are presented to adjust automatically some weights that combine those UK and U[Formula: see text] filters, independent of the dynamics of the problem. Such adaptive fuzzy hybrid unscented Kalman/[Formula: see text] filter (AFUK[Formula: see text]) is based on the combination of gain, a priori state estimation, and a priori measurement estimation. The simulation results of an inverted pendulum and a re-entry vehicle tracking problem clearly demonstrate robust and better performance of this new AFUK[Formula: see text] filter in comparison with those of both UKF and U[Formula: see text]F, appropriately. It is shown that, therefore, the new presented AFUK[Formula: see text] filter can simply eliminate the need for either UKF or U[Formula: see text] F effectively in the presence of Gaussian and/or non-Gaussian noises.