The erodibility index method proposed by Annandale was a semi-empirical methodology to measure the resistance of rock under a range of hydraulic conditions. Two modifications for determining the erodibility index were made to improve the evaluation of bedrock erosion based on the comparison of the calculated results and the field evidences. First, the block size number was modified for the estimation of RQD. Second, the original empirical function of the critical stream power estimated in terms of the erodibility index may overestimate the actual critical stream power in open channel flow. A new relationship between the critical stream power and the erodibility index was obtained through the regression of the measured erosion data. Then, the usage of the original relative ground structure number on the rate of knickpoint retreat does not justify the trend of measured rates of knickpoint retreat. The disagreement might be explained by the difference in the controlled erosion processes.