This paper presents a verification of CitySim, a largescale building energy simulation tool based on a simplified thermal model. First, the main assumptions behind the simplified model are presented. Then, CitySim is compared to other detailed simulation tools on case studies defined in the Building Energy Simulation Test (BESTEST) validation procedure. Finally, the predictions of CitySim regarding the annual heating load are compared with the monitored consumption of a building located on the campus of the Swiss Federal Institute of Technology Lausanne (EPFL). Regarding the BESTEST, CitySim gives annual heating and cooling results together with peak heating and cooling loads that are within the acceptable range defined by reference simulation tools except for the annual heating load of case 960 and the peak cooling of case 610 but the differences are rather small (less than 1%). Regarding the EPFL campus building, CitySim gives an annual heating consumption within a 5% range of the monitored consumption of two reference years. The paper concludes that despite its simplified thermal model, CitySim results remain consistent with more detailed programs and the monitored heating consumption of an office building. These results reinforce the confidence in the tool to predict annual and peak load energy needs for conditioning buildings.