The purpose of this study was to develop a SSVEP-based Brain-machine interface (BMI) rehabilitation training system for spinal cord injured persons. We designed a visual stimulation program with two checkerboard images flickering using C#. Six electrodes were attached at locations POZ, PO 3, PO4, OZ, O1 and O2, according to the international 10-20 system, and reference electrodes placed at A1 and A2. Electroencephalographic (EEG) signals were amplified 5,000 times with a digital low-pass filter with a cutoff frequency of 35 Hz and recorded at the sampling rate of 200Hz with 10-bit resolution using BIOPAC system. Fast Fourier transform (FFT) based spectrum estimation method was applied to control the functional electrical stimulation (FES) rehabilitation system. FES control signals were digitized and transferred from PC to the microcontroller using Bluetooth communication. The frequency of FES was set to 10~20 Hz, the pulse width at 200us, and the stimulation amplitude at 10~20mA. This study showed that the present SSVEP based rehabilitation training system could make successfully muscle movements, inducing electrical stimulation of forearm and lower-arm muscles in healthy volunteers.