Abstract The variable refrigerant flow system has received much attention due to its energy saving potential and design flexibility. A load responsive control of the evaporating control in the VRF system, which aims to reduce the cooling energy consumption of the VRF system, has been developed. This newly developed control consists of two algorithms. It starts with the load determinant algorithm, which evaluates the levels of internal cooling loads of individual indoor units. The second algorithm determines the target evaporating temperature based on the outcome of the first algorithm. A series of experiments in a multicalorimeter, which is composed of one outdoor unit chamber and two indoor unit chambers, indicates that increasing the evaporating temperature can reduce the electricity consumption of the VRF system by up to 35% without impairing the energy efficiency of the VRF system. The annual energy consumption of a typical office building with a VRF system has been simulated with whole-building energy simulation (EnergyPlus), and the EnergyPlus runtime language is used to make a code to model the performance variation of the VRF system as a function of the evaporating temperature. Our simulation results demonstrate that the annual cooling energy consumption is lowered by 14%.