DNA sequence analysis provides excellent evidence for the origin of new genes, encoding new enzyme specificities or isozymes, via gene-duplication. New genes which arise in this way are likely to have arisen via silent gene intermediates. Such ‘silent’ genes are conceptually distinct from ‘cryptic’ genes which may also be silent; whereas cryptic genes are thought to be retained due to periodic selection, silent genes would be expected to have only a transient existence in the genome. Only very few of the known inactive genes are possibly (and with varying degrees of likelihood) of the ‘silent’ type.