This paper investigates the recursive filtering problem with fading measurements and cyber attacks for multisensor multirate networked systems (MRNSs) under Round-Robin protocol (RRP). By exploiting the lifting technique, the sampling periods for both sensors and the state of the system are uniformed. It is assumed that the phenomenon of fading measurements which better describes practical engineering arises stochastically, and the attenuation coefficients of which are described by a set of random variables with known statistical properties. In order to fully utilize the limited communication resources, RRP is introduced in the sensor-to-filter channel. Considering the measurement outliers, a saturation function is adopted in the filter structure to suppress the anomalous innovations, so as to reduce the negative impact of the outliers. By means of matrix difference equation, an upper bound is first obtained on the filtering error covariance, and the filter gain is designed to minimize the obtained upper bound by partial derivation. Moreover, the exponential boundedness of the filtering error dynamics is analyzed in the mean square sense. Finally, a numerical simulation example is given to demonstrate the validity of the proposed recursive security filtering scheme.