In the Society of Pharmaceutical Medicine Report on Pharmacogenetics, and specifically in the section on Ethical Issues, reference was made to the Icelandic Health Sector Database (HSD).[1] It was concluded that this database was not an ideal model for such a purpose, this conclusion being based on two quite different areas of concern. One dealt with serious issues regarding consent, or the lack of it, and the other dealt with the possibility that the premise that Icelanders were genetically homogeneous, which was one reason for choosing Iceland in the first place, was false. The setting up of this database had been heavily criticized by the (Icelandic) National Bioethics Committee and Data Protection Commission, but it had nevertheless been established by an Act of Parliament. Criticisms particularly concerned issues of informed consent and of the consequences of granting a licence to a single commercial organization, deCODE. However, the greatest concerns were that anonymization, even with encryption, would be inadequate, and that the population of Iceland had been recruited into the setting up of this database without their full knowledge. Certainly the entitlement to opt out of the database was difficult to understand, let alone achieve. Other concerns related to the regular opening of the HSD for updating personal health data, which it was felt would jeopardize its confidential nature, and to the possibility of linking the HSD to other databases without consent, with consequent implications for both life insurance and employment. All these concerns remain. More seriously, and on the second area of concern mentioned above, a report has now been published from Iceland challenging, with evidence, the concept of Icelandic homogeneity.[2] This well researched article virtually demolishes the frequently publicized statement that there is a special genetic homogeneity of the Icelanders and is based on a systematic extensive reanalysis of mitochondrial DNA (mtDNA) variation by examining primary data from original sources on 26 European populations. The results actually showed that Icelanders, using two specific parameters, are amongst the most heterogeneous Europeans and, using a third parameter, that the pattern of their genetic differences has the same shape as European populations in general, and shows no evidence at all of bottlenecks of numbers in Iceland. Therefore, this is totally at variance with what is still vigorously being promulgated. How has this happened? In a commentary on Arnason’s[2] article in the same edition of the journal in which the article appears, Professor Peter Forster of the Molecular Genetics Laboratory at the University of Cambridge offers an explanation.[3] Recollecting that he himself had had been involved in mtDNA sequencing in the past, Forster described how the first sequences had contained errors that only became apparent in subsequent phylogenetic network analyses,[4] prompting him and his colleagues ever after to take great pains in proofreading before publication. It seems that errors in sequence data actually occur quite frequently, and that whereas the deCODE geneticists had argued that mtDNA in Iceland is relatively homogeneous, they had taken the sequences which led them to that conclusion from a conveniently compiled public database which, as Arnason demonstrates, is riddled with copying errors relating to the mtDNA of Danes and Germans, resulting in a qualitatively different ranking of Icelandic diversity. Does this matter? Forster convincingly argues that it does. He points out that a number of fundamental research papers, including some that have made remarkable claims for the origin of mankind or of whole continental populations[4-6] or for mtDNA recombination[7] have been criticized or rejected, respectively, due to the extent of the primary data errors which he himself, as well as others, has described and published.[8-11] What therefore seems to be happening, in a science which is still being nurtured in its infancy, is that papers are being published which are riddled with inaccuracies, only some of which are subsequently highlighted, and that this is occurring to the extent that it is throwing real doubt onto the validity of important premises. The homogeneity of Icelanders is just such a premise. Arnason is rightly seriously concerned about what he has uncovered, particularly as the immediate response from the deCODE geneticists, published in an article in Nature,[12] is that what he has revealed is not disputed but that his interpretation of his revelation is misrepresented. There is no substantiation as to the nature of this misrepresentation. Furthermore, deCODE have stated that this revelation would only have a ‘limited’ effect on Iceland’s position in the hierarchy of European genetic homogeneity. This is an arrogant, indeed dangerous, dismissal of a valid scientific reanalysis, particularly as other workers in the field of genetics have followed the example of Forster and acknowledged and corrected any errors in their original sequencing, once they have been identified.[7,13] Attitudes such as this do not help to inspire confidence in the integrity of scientific data, even when most of it is probably ethically and scientifically valid. Plans for the use of the Icelandic HSD by researchers, including those from the pharmaceutical industry, have been discussed for some time. Indeed, deCODE, being a commercial organization, would be expected to market the database. Were the database to be the subject of proper consent procedures and to be scientifically accurate, this would be fair enough. The current COMMENTARY Int J Pharm Med 2003; 17 (1): 5-6 1364-9027/03/0001-0005/$30.00/0