M.Sc. / This study compares the electricity consumption of two buildings, of similar architectural design, in Simba Office Park, Sandton, in Johannesburg. One of the buildings (Block AB) has not been retrofitted for energy efficiency, whilst the other building (Block C) is a retrofitted building. The hypothesis postulates that the retrofitted building would use less energy than the non-retrofitted one. The research methodology employed has been used internationally, as in the case reported by Levine et al. (1996), who did a study in the United States of America on retrofitting for achieving energy efficiency. Dong et al. (2005) investigated the energy savings due to the retrofitting of old Singaporean commercial office buildings. In this case, six buildings were compared before and after retrofitting, using utility bill and weather data. There have also been similar studies in China (Xu et al, 2006) and Budapest (Urge-Vorsatz & Novikova, 2008). The Budapest study also unpacked the cost of retrofitting. Using methods advocated by Probst (2004), Yalcintas (2008) and Yalcintas & Kaya (2009) for collecting data on floor space, building parameters and design, this study also collected electricity consumption data based on meter readings for the same blocks over the period March 2009 to April 2010. The retrofitting measures were documented and the associated costs noted. Interviews were conducted with key personnel such as the Central Energy Fund (CEF) House executives, the site electrical engineer, the developer, and Simba Office Park managers. Block AB had the least number of energy-efficient installations. Block C was found to be fully retrofitted, at a cost of more than R4 million. However, the energy management system, required to manage and monitor energy use, was only fully installed by November 2009. The results of this study are significant. It was found that energy consumption for Block C far exceeded that for Block AB. Thus, in this case, retrofitting did not reduce electricity consumption. The results demonstrate that in order to fully understand energy use, data collection and analysis must be ongoing. This verifies the findings of Ali (2008), Armstrong (2009) and Yalcintas & Kaya (2009) who found that we need to shift from managing buildings to managing energy use and assess and verify any recorded savings to ensure energy conservation. Computer-based building management systems play a major role in such management. Such a system was only partially in use in Block C for the duration of the study period. Thus, another finding was that the energy management system needs to be fully operational in real time, or else energy efficiencies cannot be achieved and data sets will be incomplete. This conclusion reflects the findings of Hirst (1980).