The exercise of quantifying the energy consumption data assembled through household surveys, either by the recall-based approach or the meter-based approach, remains a challenging task, especially in rural areas of developing countries. In this article, we propose a device-based bottom-up accounting method for estimating household energy consumption. This method provides microlevel disaggregated estimates at the intensive margin and documents other difficult-to-measure energy consumption such as biomass at the extensive margin. Even though measurement errors of the household survey might still exist, the structured questionnaire of daily routine behavior questions should greatly alleviate the problem. The new method supplements the existing household energy statistical system, improves its flexibility, and is particularly applicable in developing countries and/or rural areas. We apply the method to a Chinese rural household survey and discuss its differences and similarities with the conventional methods.