Accurate georegistration is required in multi-sensor data fusion, since even minor biases in spatial registration can result in large errors in the converted target geolocation. This paper addresses the problem of estimating and correcting sensor biases in target geolocation. Aiming to solve the spatial registration problem in the case where heterogeneous measurements are provided by mobile sensor (active or passive) platforms, this paper proposes a moving heterogeneous sensor registration (MDSR) algorithm based on maximum likelihood estimation. The MDSR algorithm decouples the offset biases from the attitude biases and updates heterogeneous measurements using linear minimum mean square error fusion. In particular, the MDSR algorithm is a batch algorithm that outputs estimates of the offset biases, attitude biases, and target location estimates, expressed in a common coordinate system. Calculation of the Cramér–Rao lower bound and conducting various simulation results demonstrate that the MDSR algorithm is effective and robust for moving heterogeneous sensors.