A parallel experiment has been performed on an insulated roof system installed in the Roof Thermal Research Apparatus at Oak Ridge National Laboratory. Separate sections of a 1.2 by 2.4 m test panel were evaluated for nine months using two different heat flux transducer (HFT) instrumentation packages, with the data being treated independently. The paper discusses the instrumentation packages, sensor calibration procedures, and data collection and analyses methods, highlighting the differences between the independent techniques. A comparison of the independently derived thermal performance results is presented. The comparison includes a discussion of the HFT test techniques for determining the in situ thermal performance of insulated building components. Limitations of these procedures are also detailed.