Sampling based probabilistic roadmap planners (PRM) have been successful in\nmotion planning of robots with higher degrees of freedom, but may fail to\ncapture the connectivity of the configuration space in scenarios with a\ncritical narrow passage. In this paper, we show a novel technique based on Levy\nFlights to generate key samples in the narrow regions of configuration space,\nwhich, when combined with a PRM, improves the completeness of the planner. The\ntechnique substantially improves sample quality at the expense of a minimal\nadditional computation, when compared with pure random walk based methods,\nhowever, still outperforms state of the art random bridge building method, in\nterms of number of collision calls, computational overhead and sample quality.\nThe method is robust to the changes in the parameters related to the structure\nof the narrow passage, thus giving an additional generality. A number of 2D &\n3D motion planning simulations are presented which shows the effectiveness of\nthe method.\n