In this paper, using a power series methodology we describe a design procedure applicable to analytic nonlinear plants. Our technique is a generalization of the linear H theory. In contrast to previous work on this topic ([3], [14], [15]), we now are able to explicitly incorporate a causality constraint into the theory. In fact, we show that it is possible to reduce a causal optimal design prob...