Three parameters which are critical for the development of non-toxic orally active iron chelators are identified: bioavailability, selectivity for iron (III) and distribution and toxicity. Each is discussed in detail. Arguments are presented for the use of bi- and tridentate ligands as opposed to hexadentate ligands. The discussion leads to the identification of 3-hydroxypyridin-4-ones as compounds with a unique potential for iron chelation under clinical conditions. The prodrug concept utilising efficient liver first-pass kinetics is introduced.