With　the　Semantic　Web([1])　has　been　paid　more　and　more　attention,　there　are　more　and　more　related　applications.　For　instance,　Google　has　provided　semantic　retrieval　functions,　you　can　get　the　information　of　some　famous　people　directly.　RDF[2]　database　as　the　data　storage　tool　of　Semantic　Web　also　gets　many　attentions.　Now,　there　are　some　well-known　RDF　databases　such　as　Jena([3]),　Sesame([4]),　Bigdata([5])　and　so　on.　The　role　of　RDF　database　is　to　provide　data　service　for　Semantic　Web　applications.　The　important　metric　of　database　is　the　query　efficiency.　Now,　most　of　RDF　database＇s　efficiency　has　not　reached　high　level　and　there　is　a　large　gap　between　it　and　commercial　demand([6][7]).　In　this　paper,　we　present　a　two　layer　cache　mechanism:　RC　(RDF　database　Cache)　to　speed　up　the　queries　and　relieve　the　problem　that　the　runtime　of　some　queries　is　too　long　in　RDF　database.　By　implementing　and　test,　the　query　efficiency　can　increase　thirty-one　percent　or　more　after　using　this　cache　mechanism.