The non-energy use of fossil fuels is a source of carbon dioxide (CO2) emissions that is not negligible and has been increasing substantially in the last three decades. Current emission estimates for this source category are subject to major uncertainties. One important reason is that non-energy use as published in energy statistics is not defined in a consistent manner, rendering calculation results based on these data incomparable across countries (concerns in particular the Intergovernmental Panel on Climate Change (IPCC) Reference Approach). Further reasons are the complexity and interlinkage of the energy and material flows in the chemical/petrochemical sector and the current use of storage fractions as default values in the IPCC Reference Approach, which are based on a different definition of storage and refer to other flows than those available from energy statistics. Several other shortcomings of the IPCC Reference Approach are identified in this paper, e.g. the fact that it neglects international trade of synthetic organic products. In order to improve emissions accounting, the Non-Energy Use and CO2 Emissions (NEU-CO2) network developed a model called Non-Energy Use Emission Accounting Tables (NEAT), which is based on Material Flow Analysis (MFA). The NEAT model and other MFA approaches have been applied to several countries. In this paper, the results for Italy, Japan, Korea, the Netherlands and the USA are compared with the values published in National Communications to the United Framework Convention on Climate Change (UNFCCC). It is shown that the international harmonisation of the data sources (energy statistics) and the methods appliedwould lead to substantially different emissions results for some countries, in the order of several percent. Moreover, the NEAT model and the other MFA have proved to be a valuable tool to identify errors in energy statistics. These results confirm the need for enhanced efforts to improve and harmonise energy statistics and estimation methods for CO2 emissions from non-energy use. The articles in this special issue contribute to reaching these goals.