Tunnel　boring　machine　(TBM)　performance　prediction　has　played　an　important　role　in　the　selection　of　construction　method,　project　planning,　construction　planning　and　management.　Headrace　tunnels　No.1　and　No.3　of　Jinping　II　Hydropower　station　were　excavated　respectively　by　two　tunnel　boring　machines.　On　the　base　of　the　obtained　information　of　rock　masses　conditions　and　rock　masses　environmental　conditions,　such　as　in-situ　stress　and　groundwater　condition　along　the　tunnel　alignments,　the　rock　mass　was　sectioned　according　to　the　influence　of　key　factors　on　TBM　excavation.　NTNU　and　rock　mass　characteristic　TBM　performance　prediction　models　were　used　to　predict　the　penetration　rate.　And　NTNU　cutter　life　mode　and　CAI　value　were　used　to　predict　the　cutter　wear.　By　comparing　the　results　of　in-situ　TBM　penetration　tests　with　the　prediction　penetration　rates,　the　applicability　and　errors　of　these　penetration　rate　prediction　models　were　discussed.　During　TBM　construction　process,　the　rock　mass　conditions,　TBM　operation　parameters,　TBM　operation　results　were　collected　and　analyzed.　The　actual　rock　mass　conditions　and　excavation　results　were　compared　with　the　rock　mass　conditions　obtained　for　TBM　performance　prediction　and　predicted　TBM　performance　results.　The　reasons　of　making　the　difference　were　studied.　Generally,　the　penetration　rate　prediction　is　reliable　in　most　cases.　But　in　the　condition　of　high　in-situ　stress,　these　TBM　penetration　rate　prediction　models　need　to　be　modified.　Due　to　the　influence　of　the　high　in-situ　stress　and　rock　bursts,　the　abnormal　cutter　failure　is　too　much.　Thus,　the　predicted　results　for　cutter　life　are　not　good.