Over the past century, patterns of phenotypic inheritance have been observed that are not easily rationalised by Mendel's rules of inheritance. Now that we have begun to understand more about non-DNA based, or 'epigenetic', control of phenotype at the molecular level, the idea that the transgenerational inheritance of these epigenetic states could explain non-Mendelian patterns of inheritance has become attractive. There is a growing body of evidence that abnormal epigenetic states, termed epimutations, are associated with disease in humans. For example, in several cases of colorectal cancer, epimutations have been identified that silence the human mismatch repair genes, MLH1 and MSH2. But strong evidence that the abnormal epigenetic states are primary events that occur in the absence of genetic change and are inherited across generations is still absent.