In this paper, the author does not so much try to give a blueprint for the application of sampling methods to vital registration and vital statistics as to show the opportunities for their use and the advantages to be derived from them. In the less developed areas of the world, modern sampling methods make it possible to obtain very accurate national statistics in the early stages of the establishment of a vital registration and vital statistics system and will lead to its more orderly and efficient development. In areas where more or less complete registration exists, the use of sampling may result in a reduction of costs and an improvement in the quality and currency of the data obtained.The sample vital statistics system proposed by the author should comprise complete primary registration units or combinations of them, representative of the entire universe for which statistics are wanted. The selection of these, however, must be made at random; but, in order to avoid bias, the units should be taken with probabilities proportionate to their size.After discussing the ways of carrying out his proposal and the relation of a sample vital statistics system to health programmes, the author considers the use of the sample system as a supplement to a complete system and the advantages of sampling for quality control, checking the completeness of registration, preparing advance tabulations, and conducting supplemental surveys and research.