The National Cancer Institute (NCI) sponsored clinical evaluations of investigational 'regional' hyperthermia systems at four clinical institutions. To support this project, the Center for Devices and Radiological Health (CDRH) developed a series of test instruments to evaluate the magnitude and repeatability of the induced heating by radiofrequency (RF) systems. Data from three institutions using the same model hyperthermia system have been analyzed. After heating, the average temperature from measurements taken at several points in the test phantom at each institution agree within +/- 0.002 degrees C. These differences are about equal to the measurement uncertainty. Thus, this technique can be used for preclinical evaluation and quality control of the total system operation. After one of the institutions relocated its hyperthermia system, a subsequent set of data showed inconsistencies compared to their earlier data. Investigation traced this to cable loss and power meter interference. From the analysis of the data from the three institutions, the utility of the CDRH RF phantom for hyperthermia systems evaluation is demonstrated.