The recent report of a genetic defect in mice being partially corrected using a combination of somatic cell nuclear transfer (NT) and gene therapy illustrates an important application of "therapeutic cloning" in treating disease (1). The key to success in NT lies in the ability of factors within the cytoplasm of the oocyte to interact with the donated nucleus and restore its totipotency. Central to this goal is the re-establishment of an epigenetic state comparable to that acquired during normal fertilization. Sufficient evidence exists to indicate that complete epigenetic reprogramming is rarely ever achieved following NT. As a donor nucleus is usually transferred into a non-activated oocyte, so cytoplasmic factors within the oocyte can gain direct access to the donor chromosomes when they undergo variable premature condensation. These factors may help to restore totipotency and to establish a zygotic pattern of gene expression. However, the molecular events that take place during this period are poorly understood, and so mechanisms of epigenetic reprogramming remain largely speculative. Furthermore, the term "nuclear transfer" is a misnomer, belying the fact that a significant amount of cytoplasm can accompany the nucleus during transfer. Consequently, factors present within the karyoplast may counter those within the oocyte and maintain an epigenetic state more akin to that of the donor cell. Recent observations confirm a somewhat stochastic pattern of global DNA demethylation and remethylation following NT, atypical of that occurring during normal pre-implantation development. Candidates for such an effect include the methyltransferase enzymes Dnmt3a and Dnmt3b and the somatic form of Dnmt1. Information on the methylation of specific gene sequences is limited to a few studies looking at repetitive sequence and satellite DNA, but these confirm aberrant patterns of methylation in cloned embryos. However, evidence linking patterns of global DNA methylation to aberrant gene expression is currently lacking. In contrast, recent studies of mice and sheep have demonstrated aberrant imprinted gene expression following cloning by NT and following in vitro embryo culture that is associated with a varied but common range of abnormal fetal phenotypes. In mice, embryonic stem (ES) cell NT embryos develop to term with a higher efficiency than do embryos reconstructed from somatic cell nuclei, implying that the nucleus from an ES cell 1) may require less reprogramming than that of a somatic cell; and 2) has an epigenetic state more akin to that of the early embryo. Mouse ES cells, however, can exhibit a high degree of variability in DNA methylation, leading to the aberrant expression of a number of imprinted genes and to a variety of abnormal phenotypes in the very few pups that survive to term. Our understanding of the mechanisms by which imprinted genes are regulated is rapidly evolving, but how these mechanisms are perturbed by NT is not yet understood. Sex-specific imprints are initially established in the gametes but are modified following fertilization by a number of maternally inherited cytoplasmic factors within the oocyte. As a consequence of the manner in which these oocytes are frequently derived, these factors may be rendered incapable of efficiently establishing a zygotic pattern of imprints in the nucleus. Similarly, specific differentially methylated imprinting control regions, thought to help re-establish germ line imprints lost during early embryo development, may be inappropriately methylated in the newly formed zygote, thereby compromising their ability to establish the correct epigenotype following NT for the several imprinted genes within their cluster.