The residential household becomes an important contributor to carbon emissions under rapid urbanization. This study analyzes the relationship between urban residential energy usage carbon emission and the natural environment in Xiamen, southeast China, using integrated residential data from survey questionnaires and remote sensing and GIS techniques. We first eliminate household social, economic and construction impacts by establishing a standardized family and choosing housing units constructed after the year 2003. We then use ordinary least squares regression to explain the effect of buildings' surrounding natural conditions on urban energy usage carbon emissions. The results show that urban home energy usage carbon emissions were reduced by green spaces and water bodies, after building construction and household social and economic impact factors were eliminated. Water bodies show a closer relationship with carbon emissions than green spaces. These results prove that green spaces and bodies of water are important in residential energy usage carbon emission reduction. Therefore, better green space and water planning will improve sustainable urban development. (C) 2013 Elsevier B.V. All rights reserved.