The implement of gear contact fatigue life prediction（ CFLP） is critically important for maintenance and operation of gear transmission system. A novel method for gear CFLP based on vibration data collected from the run-to-failure test of involute spur gear is proposed. The multifractal detrended fluctuation analysis（ MF-DFA） is used to analyzed vibration data. As a result,the multifractal spectrum（ MFS） are obtained and the indicator for assessing gear contact condition deterioration is designed based on the width of MFS. Subsequently,the gear CFLP model is developed based on the proposed indicator and GM（ 1,1） model. The efficiency of this model is demonstrated by using experimental results. The results show that the proposed method delivers satisfactory performances with high precision and the prediction results can be used as a reference to gear life estimation.