Muscle contractions of upper limbs induced by functional electrical stimulation (FES) is a widely used rehabilitation method for stroke patients. However, FES tends to result in rapid muscle fatigue, which greatly limits activities such as FES-assisted rehabilitation exercises. In order to minimize the effect of muscle fatigue, an intelligent Muscle Fatigue Status (MFS) and an Function Impairment Level (FIL) controlled micro-stimulator are proposed. The stimulation parameters of the FES are updated according to the FIL of stroke patients initially. During the break of stimulation, mean frequency (MNF) and median frequency (MDF) are exacted from an electromyography (EMG) signal, which is directly detected from the stimulated muscle of stroke patients. The MFS that can be generated from MNF and MDF is used to control the operation mode of micro-stimulator. This implantable micro-stimulator will be designed and fabricated using system on chip technology.