This paper deals with a new version of perturb and observe tracking algorithm for maximum power extraction from the solar photovoltaic panel, which has self-predictive and decision taking ability. The working principle of self-predictive perturb and observe (SPP&O) algorithm is based on three consecutive operating points on the power-voltage characteristic. Out of three points, first two points very smartly detects the dynamic condition, as well as in normal condition, quickly searches the maximum power point (MPP) region. Moreover, by using a circular analogy, all points decide the optimal operating position for next iteration, which is responsible for quick MPP tracking as well as improved dynamic performance. Here, in every new iteration, the step-size is reduced by 90% from the previous step-size, which provides an oscillation-free steady-state performance. The effectiveness of the proposed technique is validated by MATLAB simulation as well as tested on hardware prototype. Moreover, comparison between SPP&O algorithm and state of art methods is made. Its satisfactory dynamic and steady-state behaviors with low algorithm complexity as well as the low computational burden of the SPP&O algorithm show the superiority over state of the art methods. © 2010-2012 IEEE.