This paper is concerned with the joint decision and control strategy for a class of nonhomogeneous Markovian jump systems subject to input constraints. By assuming the mode transition rate matrix (MTRM) of Markovian jump systems is piecewise homogeneous instead of homogeneous, a higher-level Markov chain represented by the higher-level mode transition rate matrix (HMTRM) is introduced to describe the variation of MTRM. Based on this, taking into account the minimization of system cost, a novel joint decision and control strategy is proposed in the presence of input constraints. This strategy consists of two sections: decisions are applied to the HMTRM such that the occurrence probability of subsystems is adjusted ultimately; next on the basis of altered HMTRM, controller is designed in view of input constraints. In comparison with conventional control method, this strategy not only guarantees the stability of …