When houses are retrofitted with cavity wall and loft insulation, the actual energy savings derived from measurements of before and after fuel consumptions are commonly found to be considerably less than expected from the model predictions. This discrepancy, termed a ‘reduction factor’ is caused in part by changed internal temperatures, a ‘comfort factor’, with the remainder due to other causes such as insulation performance, occupancy and ventilation. This study reviewed 13 papers relating to the relationship between the predicted and the actual energy saving that occurs when houses are insulated. These papers can be divided into a) ‘strategic’ studies using simplified data from large groups of dwellings, b) detailed analyses of specific monitoring exercises with information from smaller groups of dwellings, and c) specific studies of individual measures including cavity wall insulation, loft insulation and ventilation. It was possible to quantify either a reduction factor or a comfort factor from five of the reports. These figures suggest that the best estimate for the reduction factor was 50% of the theoretical expected energy savings, with the comfort factor contributing 15% to this. Various issues were identified as relevant to the difference between the predicted and measured energy saving; these include: • The performance of thermal insulation; • Ventilation, which is a significant energy burden and is largely under occupant control, is difficult to represent in theoretical models. • Separation of energy use for hot water, cooking, lighting and other appliances from space heating consumption. • The correction of energy use for different outside temperatures before and after insulation • The use of appropriate energy models to model specific houses.