The Least Mean square (LMS) algorithm has been extensively used in many applications due to itssimplicity and robustness. In practical application of the LMS algorithm, a key parameter is the stepsize. As the step size becomes large /small, the convergence rate of the LMS algorithm will be rapidand the steady-state mean square error (MSE) will increase/decrease. Thus, the step size provides atrade off between the convergence rate and the steady-state MSE of the LMS algorithm. An intuitiveway to improve the performance of the LMS algorithm is to make the step size variable rather thanfixed, that is, choose large step size values during the initial convergence of the LMS algorithm, anduse small step size values when the system is close to its steady state, which results invariable stepsize Least Mean square (VSSLMS) algorithms. By utilizing such an approach, both a fast convergencerate and a small steady-state MSE can be obtained. Although many VSSLMS algorithmic methodsperform well under certain conditions, noise can degrade their performance and having performancesensitivity over parameter setting. In this paper, a new concept is introduced to vary the step sizebased upon evolutionary programming (VSSLMSEV) algorithm is described. It has shown that theperformance generated by this method is robust and does not require any presetting of involvedparameters in solution based upon statistical characteristics of signal.