The temperature distribution of riser reactor plays an important role in fluid catalytic cracking unit (FCCU) because it directly affects the quantity and quality of the product. However, there are only limited thermocouples installed on the FCCU riser. To monitor the axial temperature distribution of the riser, we proposed a neighborhood optimization-based distributed moving horizon estimation (N-DMHE). The riser reactor is divided into several subsystems according to the location of thermocouples, and the temperature distribution of every subsystem is estimated by an individual small-scale moving horizon estimation (MHE). These MHEs are solved serially in distributed framework. Consequently, the computational time is reduced to satisfy the requirement of online implementation. The performance index for each subsystem includes cost functions not only by itself but also by its neighbors, which is called neighborhood optimization. The neighborhood optimization is a cooperative strategy so that the global performance can be efficiently improved. The proposed method is applied to a FCCU in JiuJiang Petrochemical Co. Ltd. The experiment results show the effectiveness of the proposed method.