It has long been recognized that thermal response factors offer an accurate way to describe the characteristics and performance of a building or of its components. Although the existing techniques use time series methods they need to have carefully controlled thermal conditions in order to achieve consistent results. In this work a novel method for estimating response factors is applied to unsteady times series temperature and heat flux data collected from a test cell. The novel response factor estimation method is based on new time series analysis techniques developed by the authors. The response factor values of a brick wall in the test cell are estimated using the novel method and using a standard time series analysis method. The U-value of the wall is the area under the response factors and the U-value determined from each method is compared with the theoretical value and the value obtained using the ratio of mean values method. The method extracts response factor values directly from the measured thermodynamic field values, and can be used to analyse linear or non-linear time invariant physical processes which experience stationary stochastic boundary conditions.