The study on which this article is based is hoped to be the first of a series investigating the use of dynamic programming in agricultural allocation decisions. The model presented in this article was constructed in simple form to study once-over sequences of allocation of land to some activity of long-run benefit. The activity chosen was pasture improvement, but as will be seen later the section of the model which describes pasture improvement could be altered to make the model specific to some other relevant activity if desired. The presentation of this article is as follows: First the nature of dynamic programming as an analytical tool is explained. Then the dynamic programming relationships in the model are constructed and a simple numerical example is given to illustrate the computational procedure. The second part of the model, which derives the pasture improvement return functions, is then shown. Some general computational problems are discussed, following which the properties of both parts of the model are investigated in some detail.