Relationship of a-Silicon and TFT characteristic has been studied in SF6-Cl2 plasma at 200mtorr, the remained thickness of channel a-Silicon has been measured at different etch time, and the TFT characteristic have been measured by EPM System. Point out here, the a -Silicon of channel department from three layers of material L: a-Silicon, H: a-Silicon, and the n+ a-Silicon according to 21.74%, 56.52%, 21.74% comparison deposition. But among them, the n+ a-Silicon is the part that the channel department has to clean, rest two part is the factor which influences TFT characteristic. So need to consider the case in the whole experiment process: the percent of H: a-Silicon in all the a-Silicon, namely so-called H: a-Silicon%. Because it is not only promising the n+ a-Silicon have to be all cleaned, but also prevent the L: a-Silicon below from attacking, so time in experiment's selection has to start from a certain fixed-point. Testing the result elucidation that remained thickness of H: a-Silicon is 10.8% while the TFT characteristic becomes to differ obviously. So draw a conclusion that the TFT characteristic hardly has relationship with remained thickness of H: a-Silicon whose in the range of 25%~45% of a-Silicon, and has been greatly affected while remained thickness of H: a-Silicon lower than 25% of a-Silicon.