An increasing concern amongst geneticists is epidemiological data, as well as personal data, which is then used by insurance companies. This is an immense source of concern for several reasons, but before looking at the reasons for our concern, it would be well to look at the mode of inheritance which many of these genetic disorders follow. Those genetic conditions of which the public is most aware are often inherited in a relatively straightforward way. Among these we can count cystic fibrosis, sickle cell anaemia, Huntington’s chorea and Duchenne muscular dystrophy. The most straightforward inherited condition of these are cystic fibrosis and sickle cell anaemia; you can be an unaffected carrier with one defective gene, but one perfectly normal one which functions correctly and that is all you need. If you should marry another carrier, however, simple patterns of inheritance indicate that the probability of producing an affected child is one in four; the probability of producing a non-carrier child is also one in four. The probability of producing a carrier child is one in two, or 50%. This is a probability, nothing more, so who should have access? When it comes to Huntington’s chorea the situation starts to become a little cloudy and, with further examples which we shall look at, will become very difficult to interpret. Huntington’s chorea is generally considered to be a dominantly inherited condition. But wait, why do the onset of symptoms occur at different times of life and with different rates of progression? Surely this means that it is not just a case of dominant/recessive? But if it was, then it would be easy to define the mode of inheritance. But it is not so easy, even in this case where we can say with certainty that an individual does or does not carry the gene and therefore will or will not have the disease; we are still unable to say when the disease will manifest itself. When it comes to other forms of inheritance the story becomes ever more complicated and the questions which need to be addressed ever more convoluted. A very good example of this is found in breast cancer. It is known that breast cancer is associated with a gene called BRCA1. About 5% of breast cancers are the result of this gene, and about 70% of those women who carry the gene will die as a result before they are 70 years old. But, of course, that leaves us with 30% of carriers who never succumb to the disease. So the question is simple: what protects the lucky 30%? This is unknown and there is no way of predicting if a BRCA1 carrier is in the 70% or 30%. So it is here that the ethical consideration comes in and possibly the legal challenge. Should an insurance company be able either to insist either on an individual having a test, or having had the test voluntarily, be able to insist on seeing the results? I do not think so. No matter what the Association of British Insurers says, its logic is fundamentally flawed. There are two basic reasons for saying this, but it would be well first to look at the manner in which insurance works. When calculating insurance premiums, there are two basic ways in which it is done. The first is called mutuality. This keeps all the people at the same risk together and therefore sharing the costs and consequently an insurance company would like lots of low risk people to maximise profits. This is the situation generally used in America where health insurance for some individuals can be impossible to find because they are perceived to be in a high risk group. The