Real-time information feedback delivered via technology has been reported to produce up to 20 percent declines in residential energy consumption. There are however large differences in estimates of the effect of real-time feedback technologies on energy use. In this study, we conduct a field experiment to obtain an estimate of the impact of a real-time feedback technology. Access to feedback leads to an average reduction in household electricity consumption of 5.7 percent. Significant declines persist for up to four weeks. In examining time of day reduction effects, we find that the largest reductions were observed initially at all times of the day but as time passes, morning and evening intervals show larger reductions. We find no convincing evidence that household characteristics explain heterogeneity in our treatment effects; we examine demographics, housing characteristics and psychological variables.