This paper discusses a pragmatic randomised controlled trial of a fuel poverty intervention undertaken in NE England over a four-year period, starting in 2000/2001. Home energy efficiency was measured through Standardised Assessment Procedure (SAP) ratings in each year of the trial. The trial group received an energy efficiency intervention package in year three, and the control group in year four. Year three room temperature data for a sub-sample of 100 households were obtained. A comparison group of households not living in fuel poverty were also surveyed in all four years of the study. The intervention improved SAP ratings by 12 points, generating room temperature increases of about one degree Celsius. Families did not respond to energy efficiency gains by reducing their heating expenditure. The intervention generated improvements in satisfaction with household warmth. Its receipt was not associated with gains in self-reported health. However, modest correlations between room temperatures and better social functioning, as measured by the SF36, was found.