In building design, several approaches have been proposed for coupling computational fluid dynamics (CFD) and energy simulation (ES) to perform analyses of thermal environments. The unsteady analysis of thermal environments within buildings containing offices and colonnade spaces is difficult to perform using an ES that represents the space with a single mass point, owing to excessive predictive heat loss; therefore, CFD has typically been used instead. Although it is possible to divide the space into zones using ES, it leads to excessive predicted heat loss and the prediction of heat movement due to the influence of strong air currents, such as those associated with air conditioners. This behavior is observed because these zones are not detailed mesh divisions. To solve these problems, we proposed a method for calculating the ratio of heat contribution to zones that were pre-divided using CFD followed by the distribution of the total thermal load calculated by ES. In this study, we proposed a method for coupling ES and CFD, which enabled the unsteady analysis of a thermal environment in a large space and verified its accuracy.