Inverse analysis of rock displacement is a kind of new way of studying the parameter of physical power of rock, a lot of researchers have carried on research in this respect, and some researchers among them ask solving this question with the genetic algorithm. This paper apply master-slaver model parallel genetic algorithm (MSM-PGA ) to solve inverse analysis of rock displacement, in the course of solving, has carried on certain improvement to the original algorithm, has resolved the task of the master course and slaver course, and when concrete heredity is operated, having adopted the adaptive hereditary algorithm. The result of emulation experimental shows, MSM-PGA after improving has not merely overcome the shortcoming of original MSM-PGA, and effective improvement the speed and precision of calculation, has guaranteed the dependability of analysis result.