In　recent　years,　the　emergence　of　the　fMRI　provides　a　possibility　of　further　exploration　for　the　human　brain.　The　complex　workings　of　the　brain　are　composed　of　a　plurality　of　brain　functional　networks.　To　further　explore　the　change　affected　by　task　in　human　brain　functional　network,　the　efficiency　of　a　method　of　effective　connectivity　is　examined　by　comparing　the　two　resting　state　fMRI　data　before　and　after　a　task.　Distinguished　from　the　view　of　the　brain　network　as　a　whole　in　previous　studies,　the　method　of　Granger　causality　analysis　(GCA)　we　used　is　focused　on　the　internal　organization　within　a　brain　network.　Also,　the　method　of　functional　connectivity　is　used　to　compare　the　efficiency　of　Granger　causality　analysis.　The　results　show　that　taking　the　default　mode　network　(DMN)　as　an　example,　the　method　of　Granger　causality　analysis　can　be　more　sensitive　to　the　functional　re-organization　in　a　human　brain　network　after　a　task　compared　with　functional　connectivity.　And　the　target　of　incoming　causal　influences　area　changed　from　rPCu　to　BG　after　task.