The two important articles on melatonin in this issue make a very interesting contrast. In the first we requested and got a personal view based on years of experience and the available evidence, laced with clinical nuances and subtleties, to guide those of us who are less familiar with the field. The other, by a team who are also very experienced in the use of melatonin, has followed an evidence based approach and shows how few hard data are available when rigid criteria are applied. Surprisingly, a proposed multicentre trial in the UK has not so far been funded, despite scientific validity and ethical approval (Appleton RE, personal communication). It does not seem quite attractive enough to the unelected judges on the various charities' governing bodies. It is a very apposite example of the difficulties in proving that what we do is effective, even when most of us believe it to be so.