Path　planning　is　a　very　general　problem　in　computer　science　and　artificial　intelligence,　and　it　is　widely　used　in　many　applications,　such　as　robotics,　computer　games　and　artificial　systems.　In　many　problem　domains,　finding　an　optimal　path　will　cost　too　much　time　and　this　will　affect　the　practical　use.　During　the　decades,　many　path　planning　algorithms　have　been　proposed　to　search　for　a　bounded　sub-optimal　solution　in　order　to　improve　the　search　efficiency.　This　paper　presents　a　new　strategy　which　we　call　aggressive　heuristic　search　(or　AHS　for　short)　to　find　a　solution　instantly　and　then　find　a　bounded　sub-optimal　solution　if　time　permits.　The　AHS　includes　three　stages,　the　first　stage　is　only　using　a　heuristic　information　to　search,　which　we　call　the　exploration　stage,　and　by　this　stage　we　can　get　a　solution　instantly　and　get　a　region　that　an　optimal　path　may　pass　through,　the　second　stage　which　we　call　exploitation　stage　is　using　the　conventional　A*　algorithm　to　search　for　an　optimal　solution　in　the　region　that　given　by　stage　one,　and　the　third　stage　is　a　modification　stage　to　make　sure　that　the　final　solution　is　bounded　by　a　specific　factor　of　the　optimal　solution　by　using　the　conventional　A*　algorithm　to　search　on　the　global　region.　We　test　the　performance　of　AHS　on　2D　four-neighbor　grid　map　and　fifteen　puzzle　benchmarks,　we　find　that　when　we　need　to　find　a　solution　instantly　and　then　find　a　bounded　sub-optimal　solution　if　time　permits,　AHS　outperforms　WA*　in　the　procedure　of　generating　a　bounded　solution　and　Optimistic　search　in　the　whole　procedure　under　some　specific　bounding　factors.