This study is to develop a real-time building energy simulation method for efficient predictive control of building. One of the most critical point in predictive control is that accurate prediction should take precedence before reaching to control value. Some researches using model predictive control (MPC) based on simulations may result in error due to uncertain data input. In order to reduce error, current circumstances should be applied to the simulation; improving weather forecast data can increase the accuracy. Thus, this study focuses on developing simulation method using daily updated weather forecast data. The novel building energy simulation method proposed in this study is real-time building energy simulation in which generates a real-time weather data file. The data file consists of calculations from web-based forecast data, equations, built-in functions of EnergyPlus, and other weather elements (default) from building controls virtual test bed (BCVTB). In other words, the building energy simulation, designed in this study is an EnergyPlus-specific real-time weather data file generation method. The method generates a 24-h weather data file every day. When the weather is predicted using the proposed method, mean bias error (MBE) and coefficient of variation of root mean squared error (Cv(RMSE)) are calculated to be 1.3% and 20.1%, respectively. A case study for real-time building energy simulation is carried out, finding optimal set-point temperature of chilled water and condenser water of 8.5 degrees C and 27.75 degrees C, respectively. The implementation of the values results 23.5 kW h potential savings during the day time, compared to the predicted energy consumption. Thus, it is expected to draw more reasonable predicted control value, using the real-time building energy simulation. (C) 2016 Elsevier Ltd. All rights reserved.