This dataset contains approximate GIS (Geographic Information System) geospatial boundaries associated with each of the Grid Supply Points (GSPs) on the GB transmission network. The boundaries have been derived by generating Thiessen polygons around the lowest level grid asset for which we have a topological lookup back to GSP. In some cases, this means Thiessen polygons have been generated around Distribution Network Operator (DNO) assets (e.g. Bulk Supply Points, Primary Substations etc) and these Thiessen polygons have been aggregated up to GSP level using topological lookups provided by the DNOs, in other cases no DNO asset GIS data was available in time for the dataset release and the GSP boundary is a single Thiessen polygon. The GSP boundaries have been clipped to the DNO Licence Area boundaries, except where GIS data from DNOs provided evidence of a DNO serving demand in a neighbouring DNO Licence Area through bilateral agreement with the neighbour DNO. The GSP boundaries are used for regional modelling of the GB electricity transmission network, for example to aggregate distributed embedded generator energy flows onto the transmission network. This includes the regional FES analysis (https://www.nationalgrideso.com/future-energy/future-energy-scenarios/regionalisation-fes) and the Regional PV_Live outturn estimates published by The University of Sheffield (https://www.solar.sheffield.ac.uk/pvlive/regional/). There are multiple versions of the GIS boundaries, with older versions maintained here for posterity. The latest version of the GIS boundaries is 20220314. The 20181031 version of the dataset includes a lookup between GSPs and so-called Gnodes and Direct Connects. Gnodes are an entity used by teams within National Grid ESO for the purposes of modelling and monitoring the transmission network. In most cases there is a “ONE : ONE” relationship between “GSP : Gnode”, but occasionally a Gnode can represent several GSPs (“ONE Gnode : MANY GSPs”) and in some cases a Gnode represents an asset on the distribution network. In general, the relationship between GSPs and Gnodes is “MANY : MANY”. A Direct Connect represents demand connected directly to the transmission network. They are included in this dataset because they may theoretically host embedded generation and/or may form part of a MANY:MANY relationship with a Gnode. The 20220114 version of the dataset contains only GSPs and does not include mappings between GSPs, Gnodes and Direct Connects. Some GSPs have been grouped together because it was not possible to discern which geographical regions they each serve.