This paper addresses the problem of stabilizing a class of nonlinear systems subject to Markovian jump parameters using a robust stochastic fuzzy controller with H∞performance. The class of jump nonlinear systems is described by a fuzzy model composed of two levels, a crisp level which represents the jumps and a fuzzy level which represents the system nonlinearities. Considering the approximation error between the fuzzy model and the jump nonlinear system as norm-bounded uncertainties we develop a systematic technique to obtain a robust stochastic fuzzy controller which guarantees the L2gain of the closed-loop system in respect to external inputs to be equal to or less than a prescribed value. A simulation example on an electrical power system in co-generation scheme is presented to illustrate the approach.