A number of methods have been employed to estimate the production increases obtained from supplementary irrigation. Details of these methods and some of the disadvantages from which they suffer will be discussed later. Suffice it to say at this stage that, in the opinion of this author the methods fail to establish the fundamental relationships involved, namely, to define a production function for water, or, more basically, for soil moisture. The present paper is intended as a progress report of an attempt to relate butter production to soil moisture on three separate dairy farms. The assumption is that a functional relationship exists between soil moisture and pasture production, and, carrying this one stage further, between soil moisture and animal production. If this relationship can be defined mathematically it will be possible to estimate production responses to soil moisture changes within known limits of accuracy, an advantage which the methods referred to above do not possess.