The objective of this study is a non-invasive assessment of the thermal dose in microwave hyperthermia. We intend to monitor the induced temperature rise via the change in the resistivity of body tissue, and hence control the microwave power during treatment. An initial feasibility study using electrical impedance tomography in vivo has indicated that a microwave induced temperature difference of a suitable magnitude can be mapped from a knowledge of the temperature coefficient of conductivity for tissue (typically 2% per degrees C). The accuracy of the currently employed reconstruction algorithm is assessed from data simulated by finite element prediction methods. A circular arrangement of electrodes surrounding a conductive sheet is used to evaluate the voltage distribution on the boundary of the sheet. A range of regions inside the circular field are chosen and the conductivity is changed uniformly by steps of 1% up to 10%. Images of these changes are produced. It is noticed that the algorithm underestimates the values of resistivity change for the small areas and overestimates the change for the larger areas. We are studying results for a variety of shapes of surface regions of the body that undergo resistivity change with microwave heating applied. Further work is necessary to account for three-dimensional current paths. Preliminary results are also presented of experimental investigations of the microwave-induced temperature rise in layers of carbon-loaded paper sheet.