This document reports the results of Core Experiment (CE) 9.1.3 on complexity reduction of decoder-side motion vector refinement (DMVR). In this CE test, the existing DMVR design in the benchmark set (BMS)-1.0 is used as the base. Additionally, one early termination method is applied to terminate the DMVR process of one bi-prediction merge candidate if its L0 and L1 prediction signals are sufficiently similar. Simulation results show that compared to VTM-1.0 anchors, the proposed DMVR method provides on average {Y, U, V} BD-rate savings of {-2.60%, -2.52%, -2.64%} for RA with average encoding and decoding time of 109% and 119%.