Type footer information here Type header information here Structured Abstract [Mandatory]-Please structure in sections as shown in the author guidelines Purpose: Buildings contribute significantly to CO2 production. They are also subject to considerable taxation based on value. Analysis shows that whilst similar attributes contribute to both value and CO2 production, there is generally a loose relationship between the two. If we wish to use taxation to affect policy change (drive energy efficiency behaviour), we are unlikely to achieve this using only the current tax base (value), or by increasing the tax take off this current tax base (unlike extra taxation of cigarettes to discourage smoking, for example). Taxation of buildings on the basis of energy efficiency is hampered by the lack of current evidence of performance. This research models the now obligatory (at sale or letting) Energy Performance Certificate (EPC) data to derive an acceptable appraisal model (marked to market, being the EPC scores) and deploys this to the entire population of properties. This provides an alternative tax base with which to model the effects of a tax base switch to energy efficiency and to understand the tax incidence effects of such a policy. Design: The research employs a multiplicative hedonic approach to model energy efficiency utilising EPC holding properties in a UK jurisdiction (Northern Ireland (NI)) as the sample. This model is used to estimate discrete energy assessments for each property in the wider population, utilising attributes held in the domestic rating (property tax) database for NI (700,000+ properties). This produces a robust estimate of the EPC for every property in its current condition and its cost effective improved condition. This energy assessment based tax base is further utilised to estimate a new millage rate and property tax bill (green property tax) which is compared against the existing property tax based on value to allow tax incidence changes to be analysed. Findings: The findings show that such a policy would significantly redistribute the tax burden and would have a variety of expected and some unexpected effects. The results indicate that whilst assessing the energy performance of houses can be a complex process involving many parameters, much of the explanatory power can be achieved via a relatively small number of input variables, often already held by property tax jurisdictions. This offers the opportunity for useful housing stock modelling-such as the savings possible from power switching. The research also identifies that whilst urban areas display the expected 'heat island' effect in terms of energy consumption, urban properties are on average more efficient than suburban / rural properties. This facilitates spatial targeting of policy messages and initiatives. Research limitations/implications: Analogous with other studies, data deficiencies introduce the risk of omitted variable bias. Modelling of the energy efficiency in the sample is limited to property attributes that are availab...