In many applications of stochastic processes, there is a need not only to predict the future values of the process, but also to control the magnitude of these values. For example, a hospital administrator is interested in keeping the occupancy of his hospital as close to capacity as possible. In this article, the first and second moment structure of such a process is derived under very general conditions. These results are utilized in determining control procedures through the regulation of input to the process. The application of these procedures is illustrated in the context of a hospital for the control of occupancy.