The optimal control problem of nuclear reactor systems is solved by using the Walsh function variational synthesis technique. In this technique, the control input rate is expanded in a Walsh series with coefficients to be selected such that a general criterion functional covering all practical situations is optimized. The nuclear reactor system is assumed to be in its integral equation form, and the nonlinearity due to the product of the reactivity input and the power level output is overcome by using in the criterion functional power-times-reactivity penalty terms rather than pure reactivity ones. A numerical example is presented to illustrate the feasibility of the approach. The results are first derived for point reactors and are then extended to space-dependent ones.