The most recent methods of analysis of additive and associative effects of feeds use the continuous system. If there are two components, which are, at most, complementary, the test of linearity or curvilinearity of regression analysis allows estimating the presence of their associative effect. For more components, the Response Surface Methodology (RSM) associated to multiple regression analysis is employed. Statistical procedures are discussed for fitting a RSM to experimental data and an outline is made of the mathematical process for finding the stationary point. The specific case of a three-component system, whose sum is the whole, is examined. Examples, even in graphic format, are produced to show the application of continuous analysis to studies on the associative effects between feeds.