South Africa’s private sector - under significant pressure to become energy efficient and employ sustainability principles - has long been implementing energy saving mechanisms. Unfortunately, there seems to exist many misplaced incentives in South Africa's public sector that prevent it from embracing energy-efficient technology. With the falling cost of LED lighting and the rising cost of electricity, however, conversions are increasingly cost efficient. Effecting these changes are increasingly urgent given the national utility-imposed rolling blackouts and climate change concerns. The primary education sector is a particularly attractive test case, since money saved on utilities can be allocated to desperately needed value-adding services in schools. From a technical perspective, however, the cost-benefit of replacements and the range of options facing decision makers could be overwhelming. To assess the impact of replacing fluorescent lights with LED lights at schools in South Africa, we propose a model that draws on smart metering data, a bench-test assessment of available LED lights and tariff rates. The model was validated with field tests at three schools and used to assess the impact at seven local schools. The results show that the setup cost differs substantially from the life-cycle cost, and that buying the cheapest lights could prove to be the costliest decision over the light's life cycle. The results also show that lights contribute from 42% to 57% of electricity expenditure, and that monetary savings of 21% to 39% are achievable by replacing fluorescent tubes with the most efficient LED lighting option available.