Quantitative methods for measurement of absolute (kinetic) surface temperatures of roof surfaces from aerial thermal data are discussed. Methods of accounting for atmospheric variation, background reflectance, path radiance, and roof emissivity effects are also discussed. The quantitative method has been tested in several research programs under the sponsorship of the New York State Energy Research and Development Authority and the U.S. Depart-ment of Energy. The results of these tests, which are presented here, represent a comparison of concurrent contact and aerial roof temperature measurements on a variety of roof surfaces. These tests indicate that kinetic roof surface temperatures can be measured from wholly airborne data with a standard error of 1.0°C. The implications of these results for airborne measurement of building heat loss are presented in this paper.