Globally, about 20–30% of total energy demand is for residential use. Yet, our understanding of household energy consumption remains obscure. Due to methodological issues, conventional residential energy research has often failed to untangle the complexities of household energy use. In addition, theoretical deficiencies have led to underestimation of the complexities of households' role (vs. buildings' role) in residential energy use processes. This research hypothesizes that households have an indirect effect on energy use – through their housing choice behaviors – that can be untangled from buildings' effect. Using data from the latest Residential Energy Consumption Survey, this paper develops a Structural Equation Model to estimate the direct, indirect, and total effects of household and housing characteristics on per-capita residential energy use in the U.S. Outcomes support the research hypothesis and recommend that, when incorporating the housing choice effect, households' overall effect on energy use is larger than what has been perceived conventionally. Findings of this study provide a new approach to understanding residential energy use by highlighting the role of households and their housing choices in shaping residential energy consumption patterns. Policy makers can incorporate housing policy into energy policy and reduce residential energy consumption more effectively.