Biometrics models describe aspects of fertility in terms of mathematical functions. A number of purposes are served by these models: description and estimation, prediction, and evaluation of the effects of changes in the parameters governing the specified relationships. This paper, by reviewing reproductive models of varying complexity, attempts to illustrate their many uses in fertility, research. As mortality in the world declines, fertility becomes, more and more, the primary determinant of population growth and structure. Recognition of this situation has spurred interest in documenting levels of fertility and causes of changes in birth rates throughout history and in investigating how rates might change in the future. Fertility research, therefore, is directed toward measuring, explaining, predicting and, in some cases, influencing numbers of births and other characteristics of reproductive behavior. Biometric models have proved useful in the pursuit of each of these goals. These models are attempts to describe various aspects of fertility by means of a mathematical function whose parameters may vary among populations. Existing models differ greatly in the variables which are included as determinants of reproduction. This paper is intended as a review of models of increasing complexity in terms of inclusion of variables affecting fertility, and an analysis of both existing and possible applications. In no way is it intended as an exhaustive analytical and critical treatment, but rather as an overview of as wide a variety of examples as is possible in a brief survey. One major purpose of models is concise description of some aspect of fertility behavior. If a particular model adequately represents reality, the characteristics of a given population can be summarized by the parameters of the model as estimated from a body of data, thereby facilitating study of variation among populations or within a population over time. When a model is postulated, methods of estimating its parameters must be devised and tested for their adequacy. Both the development and comparison of procedures and the consideration of kinds of data sufficient for estimation purposes have proved fruitful. Interestingly, in a number of cases demographers have been able to devise, with the help of a model, procedures which employ data that otherwise would be considered inadequate for estimating a specified characteristic. Thus, an extremely important service performed by models is to increase our knowledge of either current or past levels of fertility. Another way of exploiting models is in predicting or forecasting fertility or population growth as extrapolations from an appropriate model viewed as theory. Nathan Keyfitz (1971) called prediction the trisecting the angle of demography because most attempts, like those of mathematicians to trisect the angle, have failed. Yet, in