In 1986, we proposed that polypeptide sequences enriched in proline (P), glutamic acid (E), serine (S) and threonine (T) target proteins for rapid destruction. For much of the past decade there were only sporadic experimental tests of the hypothesis. This situation changed markedly during the past two years with a number of papers providing strong evidence that PEST regions do, in fact, serve as proteolytic signals. Here, we briefly review the properties of PEST regions and some interesting examples of the conditional nature of such signals. Most of the article, however, focuses on recent experimental support for the hypothesis and on mechanisms responsible for the rapid degradation of proteins that contain PEST regions.