Based　on　single　directional　rapidly-exploring　random　tree　(single-RRT)　and　bi-directional　rapidly-exploring　random　tree　(bi-RRT),　the　motion　planning　of　reaching　point　movements　for　7R　robotic　manipulators　in　obstacle　environment　is　systematically　studied.　Numerical　simulations　and　physical　experiments　are　conducted　by　using　the　single-RRT　algorithm.　A　new　version　of　bi-RRT　algorithm　is　proposed　by　incorporating　the　pose　adjustments　and　the　self-motion　in　joint　space　into　the　expansion　of　target　trees.　A　general　analytical　inverse　kinematic　algorithm　for　7R　robotic　manipulators　is　used　to　get　desired　target　configurations.　In　this　way,　the　given　target　configuration　in　the　standard　bi-RRT　is　replaced　by　a　group　of　configurations　in　the　new　algorithm.　In　a　given　obstacle　environment,　the　robot　can　automatically　choose　one　of　the　appropriate　configurations　as　the　target　node　to　direct　the　expansion　of　searching　tree　in　the　most　efficient　manner.　Through　numerical　simulations,　the　superiority　of　this　algorithm　is　further　verified.　By　using　Matlab,　VC++　and　OpenGL,　a　simulation　software　is　developed　to　further　conduct　experimental　researches　on　the　motion　planning　of　reaching　point　movements　for　7R　robotic　manipulators.　©　2012　Journal　of　Mechanical　Engineering.