Two epidemiological studies carried out in the Denver area raised the hypothesis that increased levels of power-frequency magnetic fields in residences may be associated with increased incidences of childhood cancer. This hypothesis is based on the use of a surrogate method of estimating the historical level of residential magnetic fields, using the physical characteristics of the nearby distribution wiring rather than actual magnetic field measurements. However, the actual extent to which particular internal or external conductors will contribute to the residential environment, and by what mechanism, is not well understood, leading to doubts about the hypothesis. The Magnetic Field Source Characterization Program is designed to provide information about the detailed mechanisms by which magnetic fields are produced in homes. To assist in the development of this program, two recent surveys of residential magnetic fields, one in the Denver area and the other in the Tacoma/Seattle/Everett area of the U.S., are analyzed, and the results are used to suggest guidelines for the program.