Modern district heating (DH) systems are one of the most promising heat supply solutions to reach the goal of fully decarbonised energy systems. The research goal of this paper is to evaluate the impact of installation of individual metering in DH systems on energy savings and emission reduction by application of machine learning algorithms and to predict how would this particular system upgrade measure influence the energy consumption and emissions. The research is focused on DH systems in Croatia. The results showed that dominant variable is the installation rate of individual metering devices (i.e. heat cost allocators - HCAs) and, for maximum energy saving, it should strive to 100% within a building. In that case, decrease in annual specific heat consumption in average building connected to a district heating system in Croatia is expected above 40 kWh/m2. Developed regression models show that apartments with installed HCAs could achieve about 40% reduction in heat consumption, compared to the apartments without HCAs.