There are, in the broadest sense, two mechanisms by which gene expression can be extinguished in vertebrates. The first of these is based on mass action effects of positive and negative regulatory factors and is termed activation and repression; the second is independent of positive regulatory factors but is based on the history of the affected gene and is termed silencing. It can be said, again in the broadest sense, that imprinted genes, genes subject to X inactivation, and transposon promoters are subject to silencing, while the promoters of tissue-specific genes in non-expressing tissues are controlled by activation and repression. The escape of imprinted genes from silencing through unknown mechanisms can cause developmental abnormalities and can predispose to the formation of embryonal tumours. One developmental disorder caused by loss of imprinting of genes on chromosome 11p15.5 is Beckwith-Wiedemann syndrome (BWS). This syndrome has long been known to be inexplicably common in monozygotic twins; the twins are nearly always discordant for BWS, and nearly all twins are female. A loss of imprinting model based on stochastic errors in the nucleocytoplasmic trafficking of the DNA methyltransferase DNMT1, or a paternally expressed function that opposes maintenance methylation of maternally repressed growth-enhancing genes, is proposed to explain the perplexing genetics of BWS in monozygotic twins.