**ABSTRACT:** This paper presents a comparative study of three algorithms that use occupancy grid map of environments to find a path for a mobile robot from the given start to goal – the D*, Two Way D* (TWD*) and E*. All three algorithms have ability of dynamic replanning in case of changes in the environment. The D* algorithm produces a path consisted of line segments that have discrete transitions between cell edges – a multiple of 45 deg. This path is hard to follow by a mobile robot. The TWD* and E* algorithms produce more natural paths with continuous headings of the path line segments. The criteria for comparison were the path characteristics, the time of execution and the number of iterations of the algorithms’ main while loop. The algorithms were verified both by simulation and experimentally on a Pioneer 3DX mobile robot equipped with a laser range finder.