Heating patterns and temperatures are among the most important determinants of English home energy use. Consequently, building stock models, widely used in informing UK energy policy, are highly sensitive to the assumptions they make on how occupants heat their homes. This study examined heating patterns in English living rooms and compared them to model assumptions. A time series of winter spot temperature measurements was translated into statements of the heating system being on or off during weekdays and weekend days, and the heating demand temperature estimated. The analysis showed that weekdays and weekend days are far more similar than commonly assumed. Contrary to model assumptions, homes were frequently heated outside assumed heating hours and not all homes were heated at the same time or followed the same pattern. The estimated demand temperature was about 20.6 °C, and the average temperature during heating periods was about 19.5˚C, both lower than the commonly assumed 21˚C used in models. Significantly, variability between homes in demand temperature and hours of heating was substantial. The results indicate the need to revisit some assumptions made in building stock models, and to take account of variability between homes when aiming at predicting space heating demand for an individual home.