X-ray fluorescence (XRF) is rapidly becoming established as a powerful tool in the in vivo study of body composition. One particular area of interest concerns occupational health and the monitoring of heavy metal exposure. Another concerns monitoring of the efficacy of chelation of Fe in those suffering from the hereditary blood disorder b-thalassaemia. Treatment of this particular disorder, prevalent amongst peoples of Mediterranean, Middle Eastern and south Asian descent, can lead to a build-up of Fe in the body to levels which are potentially toxic. This paper begins by reviewing some important developments in biomedical applications of in vivo XRF. Efforts to enhance the ratio of measurement sensitivity to radiation dose are highlighted. Current interests of this group are focused on the monitoring of skin iron levels in subjects undergoing treatment for b-thalassaemia. In simulated skin we are seeking to obtain, for an acceptable dose, a minimum detectable level of Fe of 10 Β΅g (g skin tissue) -1 using realistic monitoring periods. The arrangement comprises a high-output tungsten anode x-ray tube, from which a collimated beam is directed on to a copper foil of thickness 0.15 mm. This K-edge filtering gives a peak output energy of approximately 8.4 keV and a spread of approximately 1.5 keV at the FWTM. The effective 8.4 keV emission is used to excite Fe in simulated skin. The energy of the filtered source is just above the Fe absorption edge and hence close to the maximum value of the photoelectric cross-section. Preliminary results from this study indicate that for counting periods of the order of 1000 s one can detect levels of less than 10 Β΅g (g skin tissue) -1 using an entrance surface dose-rate of a few mGy min -1 .