The　efficient　knowledge　sharing　is　the　key　element　for　innovation　network＇s　success.　The　strategies　are　classified　as　＇sharing＇,　＇no-sharing＇　and　＇retaliation＇.　The　game　model　of　knowledge　sharing　is　set　up　in　terms　of　the　evolutionary　game　theory.　The　simulation　model　is　constructed　with　the　Netlogo　software　platform　which　is　efficient　in　multi-agent　simulation.　By　modeling　and　simulation,　the　evolutionary　game　equilibrium　status　of　knowledge　sharing　is　analyzed　according　to　different　payoff　value.　The　result　demonstrates　that　communication,　trust　and　punishment　are　all　important　governance　mechanisms　for　knowledge　sharing　in　innovation　network.　©　2011　IEEE.