The transition towards net zero in the United Kingdom, includes a particular focus on decarbonising heat demand in domestic buildings, due to ageing energy inefficient building stock and the large dependence on fossil fuel systems. Consequently, the UK aims to reduce space heating loads and provide it from clean sources, as well as ensure homes are warm and bills are affordable. Social housing is a primary candidate for decarbonisation given its public ownership and the vulnerability of its occupants; local authorities need to retrofit their building stock whilst assuring energy efficiency and affordability of comfort. For home energy upgrade programmes to succeed, it is necessary to understand the characteristics of current and future heating demand and identify what assumptions should be re-evaluated. Thus, this work sets out to evaluate the current state of knowledge on heat demand in social housing, and specifically aims to: (i) identify evidence gaps in the prediction of heating demand in domestic buildings in the United Kingdom, (ii) map current knowledge on heating related occupant behaviour (ii) evaluate methodologies and assumptions for estimating domestic heat demand and develop evidence-based suggestions to improve them and, (iv) add empirical evidence on occupant behaviour and their impact in heating demand within the context of social housing. First, a systematic literature review is performed to addresses objectives (i) and (ii). This review highlights lack of large-scale data collection on occupancy and heating, and insufficient evidence on the impact of smart heating controls. Secondly, in relation to objective (iii), a mapping of household typologies through the English Housing Survey 2014-15 sample suggests that current models are not representative. Results show that the most frequently household typologies used in UK building simulation, (a) a family with dependent children where the parents work full time; and (b) a retired elderly couple who spend most of their time indoors, amount to only 19% of England’s households. A more representative selection of household typologies is identified, and occupancy patterns of each group are generated using 2015 UK Time Use Survey diaries. Furthermore, to address objective (iv), the heating demand in two social housing case studies in the South of England is evaluated through the exploration of: heat billing records from 462 dwellings from a tower block complex, and setpoint records from smart thermostatic radiator valves in 47 flats in a care home. The first discovers a very low heating demand and three main types of residents: households that do not use space heating (11%), irregular households, where the transition towards the heating season is not identifiable (33%), and households with marked seasonal thresholds (56%). behaviour. The second shows evidence on distinct heating strategies and types of interactions with heating controls, as well as poor understanding of the heating systems (only 50% of residents showed a behaviour consistent with the principles of operation of the system). Overall, this analysis finds distinct user profiles and heating strategies, highlighting the variability of heat demand and the contrast between expected and actual usage of home heat controls. These findings underline the importance of moving towards a more disaggregated approach in energy modelling and have a direct application in bottom-up models, thermal comfort, and compliance assessment. The heterogeneity of heating demand calls for fit-for-purpose large scale data collection on occupant behaviour and heating, to update national standards and forecasts, as well as tailor local interventions to promote energy efficiency in homes. Defining the pathways towards net zero homes and low carbon heat supply requires understanding residents, and the potential impact of measures in terms of carbon, comfort, and health. In the current context of energy crisis and devolution, this work is highly relevant for local authorities who play a key role in assuring buildings are livable, and affordable.